WEBVTT 00:00:00.000 --> 00:00:00.964 align:middle line:90% 00:00:00.964 --> 00:00:02.410 align:middle line:90% [SQUEAKING] 00:00:02.410 --> 00:00:03.374 align:middle line:90% [RUSTLING] 00:00:03.374 --> 00:00:05.784 align:middle line:90% [CLICKING] 00:00:05.784 --> 00:00:12.215 align:middle line:90% 00:00:12.215 --> 00:00:13.340 align:middle line:90% NORVIN RICHARDS: All right. 00:00:13.340 --> 00:00:16.325 align:middle line:90% Let's progress, then. 00:00:16.325 --> 00:00:19.340 align:middle line:90% 00:00:19.340 --> 00:00:23.750 align:middle line:84% So last time we got started drawing trees for sentences. 00:00:23.750 --> 00:00:26.960 align:middle line:84% I spent a lot of time with verb phrases 00:00:26.960 --> 00:00:30.500 align:middle line:84% drawing trees for more and more elaborate verb phrases 00:00:30.500 --> 00:00:34.370 align:middle line:84% and ignoring repeated requests that I go ahead and diagram 00:00:34.370 --> 00:00:38.610 align:middle line:84% the rest of the sentence already, and I finally did. 00:00:38.610 --> 00:00:40.402 align:middle line:90% So we had sentences like-- 00:00:40.402 --> 00:00:42.860 align:middle line:84% I think this was not one of them, but this is like the ones 00:00:42.860 --> 00:00:43.818 align:middle line:90% we were talking about-- 00:00:43.818 --> 00:00:45.700 align:middle line:90% "She will tickle the child." 00:00:45.700 --> 00:00:49.865 align:middle line:84% I said we're going to do the construction of sentences 00:00:49.865 --> 00:00:51.740 align:middle line:84% the same way we did the construction of words 00:00:51.740 --> 00:00:53.360 align:middle line:84% back when we were doing morphology. 00:00:53.360 --> 00:00:55.100 align:middle line:84% We're going to use this operation "merge" 00:00:55.100 --> 00:00:57.290 align:middle line:84% which takes pairs of things and puts them together 00:00:57.290 --> 00:00:58.470 align:middle line:90% to make new things. 00:00:58.470 --> 00:01:02.600 align:middle line:84% So we'll merge "the" and "child" and get a new thing, which 00:01:02.600 --> 00:01:04.730 align:middle line:84% we'll name a noun phrase after the fact 00:01:04.730 --> 00:01:07.310 align:middle line:84% that it crucially contains a noun. 00:01:07.310 --> 00:01:10.340 align:middle line:84% And we'll merge that noun phrase with the verb 00:01:10.340 --> 00:01:12.410 align:middle line:84% "tickle" and that's how we get the verb phrase. 00:01:12.410 --> 00:01:14.070 align:middle line:84% And then we have these two other words 00:01:14.070 --> 00:01:16.250 align:middle line:90% in here, "she" and "will." 00:01:16.250 --> 00:01:19.370 align:middle line:84% And "she" is a noun, a special kind of noun. 00:01:19.370 --> 00:01:20.320 align:middle line:90% It's a pronoun. 00:01:20.320 --> 00:01:24.740 align:middle line:84% And "will," I made up a name for "will." 00:01:24.740 --> 00:01:26.180 align:middle line:90% I called it a "T." 00:01:26.180 --> 00:01:27.933 align:middle line:90% T for tense. 00:01:27.933 --> 00:01:29.600 align:middle line:84% Actually, that gives me too much credit. 00:01:29.600 --> 00:01:31.490 align:middle line:84% Other people gave it that name, I'm just 00:01:31.490 --> 00:01:33.330 align:middle line:90% transmitting to you that name. 00:01:33.330 --> 00:01:33.830 align:middle line:90% Yeah. 00:01:33.830 --> 00:01:36.260 align:middle line:84% And then we said, if we want to make 00:01:36.260 --> 00:01:38.360 align:middle line:84% a tree for the whole sentence, what we'll do 00:01:38.360 --> 00:01:42.300 align:middle line:84% is first we'll merge T with the verb phrase. 00:01:42.300 --> 00:01:44.180 align:middle line:84% So we'll merge those two things and we'll 00:01:44.180 --> 00:01:47.390 align:middle line:84% give the resulting object the label T. 00:01:47.390 --> 00:01:50.540 align:middle line:84% And then we'll merge that new object that we've 00:01:50.540 --> 00:01:53.840 align:middle line:90% created with the noun. 00:01:53.840 --> 00:01:57.560 align:middle line:84% And since, maybe you remember, "phrase" is just 00:01:57.560 --> 00:02:01.670 align:middle line:84% a name for the largest thing with a given label, 00:02:01.670 --> 00:02:06.200 align:middle line:84% that daughter of TP that we've got there, "she"-- 00:02:06.200 --> 00:02:09.080 align:middle line:84% yeah, it's a noun, but it's now also a noun phrase 00:02:09.080 --> 00:02:12.800 align:middle line:84% because it's the highest thing with its label. 00:02:12.800 --> 00:02:16.550 align:middle line:84% That noun phrase and its sister, that T bar, 00:02:16.550 --> 00:02:20.180 align:middle line:84% become the daughters of this new node that we've created, TP. 00:02:20.180 --> 00:02:23.570 align:middle line:84% I said that's the way we're going to build sentences. 00:02:23.570 --> 00:02:25.730 align:middle line:84% Immediately several of you objected. 00:02:25.730 --> 00:02:28.040 align:middle line:84% Kateryna here, for example, ruthlessly 00:02:28.040 --> 00:02:31.430 align:middle line:84% demanded that I explain why we were doing it 00:02:31.430 --> 00:02:35.270 align:middle line:84% in that order and not, for example, like this. 00:02:35.270 --> 00:02:38.180 align:middle line:84% We've got "tickle the child" just like before, but this time 00:02:38.180 --> 00:02:41.390 align:middle line:84% we'll merge "she" with "will" first. 00:02:41.390 --> 00:02:45.080 align:middle line:84% It'd be a projecting T. And then we'll merge the result of that 00:02:45.080 --> 00:02:48.330 align:middle line:84% with "tickle the child," giving you a TP again-- 00:02:48.330 --> 00:02:49.010 align:middle line:90% [COUGHS] 00:02:49.010 --> 00:02:49.970 align:middle line:90% Excuse me. 00:02:49.970 --> 00:02:51.230 align:middle line:90% Giving you a TP again-- 00:02:51.230 --> 00:02:53.840 align:middle line:90% that didn't help at all, did it? 00:02:53.840 --> 00:02:56.565 align:middle line:84% But with a different constituent structure. 00:02:56.565 --> 00:02:58.940 align:middle line:84% Several of you wanted to know, why are we not doing that? 00:02:58.940 --> 00:02:59.750 align:middle line:90% Yeah. 00:02:59.750 --> 00:03:03.530 align:middle line:84% And I attempted to convince you not 00:03:03.530 --> 00:03:07.100 align:middle line:84% to do that with the tools that we have available, 00:03:07.100 --> 00:03:11.360 align:middle line:84% the constituent structure tests that we've 00:03:11.360 --> 00:03:13.730 align:middle line:84% developed in like a day and a half of syntax. 00:03:13.730 --> 00:03:16.550 align:middle line:84% And I was not able to do that because I had not 00:03:16.550 --> 00:03:18.740 align:middle line:84% shown you sufficiently sophisticated constituency 00:03:18.740 --> 00:03:20.120 align:middle line:90% tests. 00:03:20.120 --> 00:03:25.070 align:middle line:84% So I began on the board showing you another constituency test, 00:03:25.070 --> 00:03:28.580 align:middle line:84% and I promised you that today I would use that test 00:03:28.580 --> 00:03:30.770 align:middle line:84% to demonstrate that we need the tree on the left 00:03:30.770 --> 00:03:32.780 align:middle line:84% and not the tree on the right, and that 00:03:32.780 --> 00:03:34.830 align:middle line:84% was a dramatic cliffhanger last time. 00:03:34.830 --> 00:03:36.513 align:middle line:84% And then I attempted to distract you 00:03:36.513 --> 00:03:38.930 align:middle line:84% from this by talking about other stuff for a little while. 00:03:38.930 --> 00:03:40.880 align:middle line:84% That's more or less where we were last time. 00:03:40.880 --> 00:03:42.380 align:middle line:90% Yeah? 00:03:42.380 --> 00:03:42.880 align:middle line:90% OK. 00:03:42.880 --> 00:03:45.983 align:middle line:84% So now let me attempt to distinctly show you 00:03:45.983 --> 00:03:47.650 align:middle line:84% that we should have the tree on the left 00:03:47.650 --> 00:03:48.310 align:middle line:90% and the tree on the right. 00:03:48.310 --> 00:03:49.935 align:middle line:84% And I should tell you that although I'm 00:03:49.935 --> 00:03:53.740 align:middle line:84% going to give you a piece of evidence for that today, 00:03:53.740 --> 00:03:57.430 align:middle line:84% the evidence is going to come in in bits 00:03:57.430 --> 00:04:00.010 align:middle line:84% over our discussion of syntax, which is 00:04:00.010 --> 00:04:01.900 align:middle line:90% going to last a little while. 00:04:01.900 --> 00:04:04.490 align:middle line:84% So I'll show you one piece of evidence now 00:04:04.490 --> 00:04:07.180 align:middle line:84% and if you don't like that piece of evidence, well, just wait. 00:04:07.180 --> 00:04:08.720 align:middle line:90% There will be more. 00:04:08.720 --> 00:04:11.310 align:middle line:84% But what I'm going to do today, thanks to Kateryna, 00:04:11.310 --> 00:04:13.810 align:middle line:84% is rush something that I was planning to tell you about much 00:04:13.810 --> 00:04:15.730 align:middle line:90% later. 00:04:15.730 --> 00:04:18.600 align:middle line:84% So if you don't like what I'm about to do, blame her. 00:04:18.600 --> 00:04:20.140 align:middle line:90% This was all her idea. 00:04:20.140 --> 00:04:21.220 align:middle line:90% OK? 00:04:21.220 --> 00:04:22.450 align:middle line:90% Yeah? 00:04:22.450 --> 00:04:24.780 align:middle line:90% OK. 00:04:24.780 --> 00:04:28.650 align:middle line:84% So here are some data that we talked about last time. 00:04:28.650 --> 00:04:32.910 align:middle line:84% Different sentences, I think, but the same general idea. 00:04:32.910 --> 00:04:33.580 align:middle line:90% Observation. 00:04:33.580 --> 00:04:35.455 align:middle line:84% These are both grammatical English sentences: 00:04:35.455 --> 00:04:38.580 align:middle line:84% "She would recognize Mary" and "Anyone who knows 00:04:38.580 --> 00:04:39.720 align:middle line:90% her would recognize Mary." 00:04:39.720 --> 00:04:41.400 align:middle line:90% They're both English sentences. 00:04:41.400 --> 00:04:44.340 align:middle line:84% But there's an interesting difference between them. 00:04:44.340 --> 00:04:49.050 align:middle line:84% They both contain pronouns and they contain the name Mary, 00:04:49.050 --> 00:04:52.620 align:middle line:84% but they differ in that the first sentence, "she" 00:04:52.620 --> 00:04:54.720 align:middle line:90% cannot be Mary. 00:04:54.720 --> 00:04:58.050 align:middle line:84% It has to be some other person in that first sentence. 00:04:58.050 --> 00:05:03.730 align:middle line:84% So that first sentence has to involve two different people-- 00:05:03.730 --> 00:05:06.880 align:middle line:84% one of them female, and Mary is probably also female, though 00:05:06.880 --> 00:05:07.380 align:middle line:90% who knows. 00:05:07.380 --> 00:05:08.010 align:middle line:90% Maybe not. 00:05:08.010 --> 00:05:10.850 align:middle line:90% Yeah? 00:05:10.850 --> 00:05:13.310 align:middle line:84% On the other hand, in the second sentence, 00:05:13.310 --> 00:05:15.260 align:middle line:84% "Anyone who knows her would recognize Mary," 00:05:15.260 --> 00:05:18.545 align:middle line:84% it's possible that "her" refers to Mary. 00:05:18.545 --> 00:05:21.232 align:middle line:84% Those were the data that we collectively 00:05:21.232 --> 00:05:22.940 align:middle line:84% decided on last time when we were talking 00:05:22.940 --> 00:05:24.060 align:middle line:90% about examples like this. 00:05:24.060 --> 00:05:27.320 align:middle line:84% And I showed you the second kind of example partly-- 00:05:27.320 --> 00:05:31.640 align:middle line:84% I think somebody hopefully offered me 00:05:31.640 --> 00:05:33.483 align:middle line:84% a theory of why the first one was bad. 00:05:33.483 --> 00:05:35.900 align:middle line:84% It was the theory I was hoping they would offer where they 00:05:35.900 --> 00:05:38.353 align:middle line:84% said, yeah, maybe what's wrong with "She would recognize 00:05:38.353 --> 00:05:40.520 align:middle line:84% Mary"-- what's wrong with having "She" refer to Mary 00:05:40.520 --> 00:05:43.460 align:middle line:84% is that Mary is later in the sentence. 00:05:43.460 --> 00:05:45.590 align:middle line:84% Pronouns refer to other people, but they 00:05:45.590 --> 00:05:47.840 align:middle line:84% have to be people who we've already been talking about 00:05:47.840 --> 00:05:48.798 align:middle line:90% or something like that. 00:05:48.798 --> 00:05:51.620 align:middle line:84% Totally sensible, reasonable theory. 00:05:51.620 --> 00:05:54.980 align:middle line:84% You might expect it to work that way, and it doesn't. 00:05:54.980 --> 00:05:57.710 align:middle line:84% The second sentence shows that it's possible for pronouns 00:05:57.710 --> 00:05:59.840 align:middle line:84% to refer to things that are later in the sentence. 00:05:59.840 --> 00:06:01.040 align:middle line:90% Yeah? 00:06:01.040 --> 00:06:01.910 align:middle line:90% Cool. 00:06:01.910 --> 00:06:04.273 align:middle line:84% So how are we going to distinguish these two theories 00:06:04.273 --> 00:06:04.940 align:middle line:90% from each other? 00:06:04.940 --> 00:06:06.398 align:middle line:84% That was the cliffhanger last time. 00:06:06.398 --> 00:06:13.178 align:middle line:84% I said we can distinguish these theories or these examples. 00:06:13.178 --> 00:06:14.970 align:middle line:84% So that's just the data I just ran through. 00:06:14.970 --> 00:06:16.760 align:middle line:84% "She" is not Mary in the first sentence, 00:06:16.760 --> 00:06:18.460 align:middle line:84% but "she" can be Mary in the second sentence. 00:06:18.460 --> 00:06:19.877 align:middle line:84% It doesn't have to be, but it can. 00:06:19.877 --> 00:06:22.100 align:middle line:90% 00:06:22.100 --> 00:06:22.850 align:middle line:90% All right. 00:06:22.850 --> 00:06:26.630 align:middle line:84% Now, I'm going to show you how we'll distinguish 00:06:26.630 --> 00:06:31.970 align:middle line:84% those kinds of examples using the tree that I was hoping 00:06:31.970 --> 00:06:36.710 align:middle line:84% you would just unthinkingly accept when I built it for you. 00:06:36.710 --> 00:06:39.470 align:middle line:84% And then what I'll do is show you 00:06:39.470 --> 00:06:41.660 align:middle line:84% that the theory that I'm going to show you 00:06:41.660 --> 00:06:44.180 align:middle line:84% for this pair of sentences wouldn't work if we had 00:06:44.180 --> 00:06:45.440 align:middle line:90% the other kind of structure. 00:06:45.440 --> 00:06:47.120 align:middle line:84% And then if you want, we can talk 00:06:47.120 --> 00:06:49.100 align:middle line:84% some about whether there is another way 00:06:49.100 --> 00:06:50.850 align:middle line:84% to talk about the other kind of structure. 00:06:50.850 --> 00:06:54.770 align:middle line:84% Now, my abilities as a graphic designer-- 00:06:54.770 --> 00:06:57.560 align:middle line:84% you may have picked up on this seeing my slides-- 00:06:57.560 --> 00:07:00.470 align:middle line:90% are limited. 00:07:00.470 --> 00:07:02.780 align:middle line:84% So I was not able to come up with a way 00:07:02.780 --> 00:07:05.120 align:middle line:84% to put all of the trees that you might like 00:07:05.120 --> 00:07:08.310 align:middle line:84% to be able to look at all at once on a single slide, 00:07:08.310 --> 00:07:13.190 align:middle line:84% so I'm going to put some of them up here on the board instead. 00:07:13.190 --> 00:07:19.470 align:middle line:84% We'll see if we can make reference to them. 00:07:19.470 --> 00:07:25.100 align:middle line:84% So she would-- here's the tree that people were 00:07:25.100 --> 00:07:26.840 align:middle line:90% offering as an alternative. 00:07:26.840 --> 00:07:30.800 align:middle line:90% 00:07:30.800 --> 00:07:31.340 align:middle line:90% Yeah. 00:07:31.340 --> 00:07:32.570 align:middle line:84% This is the kind of tree that people were 00:07:32.570 --> 00:07:33.540 align:middle line:90% offering as an alternative. 00:07:33.540 --> 00:07:34.940 align:middle line:84% So the tree that's on the slide is the tree 00:07:34.940 --> 00:07:36.500 align:middle line:90% that I want you to believe in. 00:07:36.500 --> 00:07:37.760 align:middle line:84% The tree that's on the board is the tree 00:07:37.760 --> 00:07:40.050 align:middle line:84% that people said, "Hey, why are we not doing it that way?" 00:07:40.050 --> 00:07:40.550 align:middle line:90% Right? 00:07:40.550 --> 00:07:44.290 align:middle line:90% 00:07:44.290 --> 00:07:46.960 align:middle line:84% I always feel nervous when I write things on the board 00:07:46.960 --> 00:07:49.340 align:middle line:90% that I want you not to believe. 00:07:49.340 --> 00:07:56.440 align:middle line:84% So just so we're clear, this is wrong. 00:07:56.440 --> 00:07:58.330 align:middle line:90% Not the right tree. 00:07:58.330 --> 00:07:59.530 align:middle line:90% That's the right tree, yeah? 00:07:59.530 --> 00:08:01.790 align:middle line:84% But I'm now going to try to convince you of that. 00:08:01.790 --> 00:08:03.580 align:middle line:84% So the reason it makes me nervous 00:08:03.580 --> 00:08:08.212 align:middle line:84% is that I know it's morning, loosely construed. 00:08:08.212 --> 00:08:10.420 align:middle line:84% It's late enough in the semester that you're probably 00:08:10.420 --> 00:08:11.205 align:middle line:90% sleep deprived. 00:08:11.205 --> 00:08:13.330 align:middle line:84% It's easy to think, well, if the professor wrote it 00:08:13.330 --> 00:08:14.920 align:middle line:84% on the board it must be true, yeah? 00:08:14.920 --> 00:08:16.923 align:middle line:90% This is not true. 00:08:16.923 --> 00:08:18.340 align:middle line:84% At least, I don't think it's true. 00:08:18.340 --> 00:08:19.382 align:middle line:90% I'll try to convince you. 00:08:19.382 --> 00:08:22.360 align:middle line:90% OK, so that tree up there. 00:08:22.360 --> 00:08:27.880 align:middle line:84% If we use that tree up there, we can 00:08:27.880 --> 00:08:30.340 align:middle line:84% account for the contrast between the two sentences 00:08:30.340 --> 00:08:31.630 align:middle line:90% on the slide-- 00:08:31.630 --> 00:08:32.590 align:middle line:90% [COUGH] 00:08:32.590 --> 00:08:34.240 align:middle line:90% My gracious. 00:08:34.240 --> 00:08:37.059 align:middle line:84% We can account for the contrast between the two sentences 00:08:37.059 --> 00:08:39.490 align:middle line:84% on the slide with the following principle. 00:08:39.490 --> 00:08:40.990 align:middle line:90% It goes like this. 00:08:40.990 --> 00:08:45.880 align:middle line:84% If you have an NP like "she" that merges with another node-- 00:08:45.880 --> 00:08:47.680 align:middle line:90% let's call it alpha-- 00:08:47.680 --> 00:08:52.510 align:middle line:84% the NP can't refer to any of the names that are inside alpha, 00:08:52.510 --> 00:08:55.550 align:middle line:84% any of the names that are dominated by alpha. 00:08:55.550 --> 00:08:59.110 align:middle line:84% I'll keep it specific to names for now. 00:08:59.110 --> 00:09:01.790 align:middle line:84% We're going to talk much more about principles like this 00:09:01.790 --> 00:09:02.290 align:middle line:90% later. 00:09:02.290 --> 00:09:05.140 align:middle line:84% We're going to develop a theory of what kinds of things 00:09:05.140 --> 00:09:08.320 align:middle line:84% pronouns can refer to and what kinds of things they can't. 00:09:08.320 --> 00:09:12.680 align:middle line:84% But this works for the pair of examples that we have here. 00:09:12.680 --> 00:09:15.610 align:middle line:84% So what's wrong with this example-- 00:09:15.610 --> 00:09:18.540 align:middle line:84% why can't "she" refer to Mary in this example? 00:09:18.540 --> 00:09:23.170 align:middle line:84% The theory would be it's because you merge "she" with that T bar 00:09:23.170 --> 00:09:26.910 align:middle line:84% there, and that T bar dominates "Mary." 00:09:26.910 --> 00:09:29.290 align:middle line:84% And therefore, this principle says 00:09:29.290 --> 00:09:30.880 align:middle line:84% "she" can't refer to anything that's 00:09:30.880 --> 00:09:34.270 align:middle line:84% inside that T bar and "Mary" is inside that T bar. 00:09:34.270 --> 00:09:37.318 align:middle line:84% That's what this principle is meant to do. 00:09:37.318 --> 00:09:39.610 align:middle line:84% Let me show you that that distinguishes between the two 00:09:39.610 --> 00:09:40.610 align:middle line:90% examples. 00:09:40.610 --> 00:09:45.970 align:middle line:84% So if instead we had "anyone who knows her"-- 00:09:45.970 --> 00:09:48.280 align:middle line:84% if we had "anyone who knows her" as the subject, 00:09:48.280 --> 00:09:50.260 align:middle line:84% "Anyone who knows her would recognize Mary"-- 00:09:50.260 --> 00:09:54.130 align:middle line:84% well, I'm going to try to get away without actually telling 00:09:54.130 --> 00:09:57.850 align:middle line:84% you how we're going to diagram all of "anyone who knows her." 00:09:57.850 --> 00:10:01.390 align:middle line:84% But "anyone" is the noun in "anyone who knows her," 00:10:01.390 --> 00:10:03.580 align:middle line:84% and it's a noun that's modified by what's 00:10:03.580 --> 00:10:05.260 align:middle line:90% called a relative clause. 00:10:05.260 --> 00:10:09.700 align:middle line:84% So "who knows her" is a kind of clause that modifies nouns. 00:10:09.700 --> 00:10:12.563 align:middle line:84% And I have given it the label question mark, question mark 00:10:12.563 --> 00:10:14.605 align:middle line:84% because someday we'll talk about relative clauses 00:10:14.605 --> 00:10:16.330 align:middle line:84% and talk about what label they ought to have 00:10:16.330 --> 00:10:17.390 align:middle line:90% and all that good stuff. 00:10:17.390 --> 00:10:19.570 align:middle line:90% But for now, it's a thing. 00:10:19.570 --> 00:10:20.350 align:middle line:90% Yeah. 00:10:20.350 --> 00:10:23.140 align:middle line:84% And what's inside it is that pronoun, 00:10:23.140 --> 00:10:26.410 align:middle line:90% "her," which got merged. 00:10:26.410 --> 00:10:29.620 align:middle line:84% Somewhere inside that relative clause is a verb phrase, 00:10:29.620 --> 00:10:33.460 align:middle line:84% "knows," and the noun phrase, "her." 00:10:33.460 --> 00:10:37.030 align:middle line:84% So if we ask, what was this noun first merged with? 00:10:37.030 --> 00:10:39.640 align:middle line:84% Well, it definitely wasn't merged with T bar. 00:10:39.640 --> 00:10:42.250 align:middle line:84% It was merged with something inside the relative clause. 00:10:42.250 --> 00:10:43.880 align:middle line:90% It was merged with this verb. 00:10:43.880 --> 00:10:44.960 align:middle line:90% Yeah. 00:10:44.960 --> 00:10:45.950 align:middle line:90% Yeah? 00:10:45.950 --> 00:10:49.190 align:middle line:84% So this principle that says, if a noun phrase merges 00:10:49.190 --> 00:10:51.470 align:middle line:84% with another node, alpha, the noun phrase 00:10:51.470 --> 00:10:54.740 align:middle line:84% can't refer to any names that alpha dominates. 00:10:54.740 --> 00:10:58.790 align:middle line:84% That distinguishes between "She would recognize Mary," where 00:10:58.790 --> 00:11:02.060 align:middle line:84% "she" is merged with something that contains "Mary," 00:11:02.060 --> 00:11:04.430 align:middle line:84% and "Anyone who knows her would recognize 00:11:04.430 --> 00:11:07.370 align:middle line:84% Mary," where "her" is not merged with something 00:11:07.370 --> 00:11:09.920 align:middle line:90% that contains "Mary." 00:11:09.920 --> 00:11:12.440 align:middle line:84% Do people see that this principle draws the distinction 00:11:12.440 --> 00:11:15.870 align:middle line:90% that we wanted to draw? 00:11:15.870 --> 00:11:16.800 align:middle line:90% OK. 00:11:16.800 --> 00:11:17.700 align:middle line:90% Giray? 00:11:17.700 --> 00:11:20.700 align:middle line:84% AUDIENCE: [INAUDIBLE] somewhere in the subtree 00:11:20.700 --> 00:11:21.948 align:middle line:90% or does it have [INAUDIBLE]? 00:11:21.948 --> 00:11:24.240 align:middle line:84% NORVIN RICHARDS: No, that's more or less what it means. 00:11:24.240 --> 00:11:28.820 align:middle line:84% So trees-- you think of the basic relation in trees 00:11:28.820 --> 00:11:31.100 align:middle line:84% as "immediately dominate," so a node immediately 00:11:31.100 --> 00:11:33.770 align:middle line:84% dominates the nodes that were merged to create it. 00:11:33.770 --> 00:11:35.870 align:middle line:84% And dominate is the transitive closure 00:11:35.870 --> 00:11:37.050 align:middle line:90% of immediately dominate. 00:11:37.050 --> 00:11:39.178 align:middle line:84% So you dominate the things you immediately dominate 00:11:39.178 --> 00:11:41.220 align:middle line:84% and the things those things immediately dominate, 00:11:41.220 --> 00:11:43.803 align:middle line:84% and so on until you run out of immediate domination relations. 00:11:43.803 --> 00:11:45.740 align:middle line:90% That's what "dominate" means. 00:11:45.740 --> 00:11:47.660 align:middle line:84% To put it more graphically, you dominate 00:11:47.660 --> 00:11:52.310 align:middle line:84% the things that are related to you by lines pointing down. 00:11:52.310 --> 00:11:54.260 align:middle line:90% That's what this means. 00:11:54.260 --> 00:11:54.890 align:middle line:90% Raquel? 00:11:54.890 --> 00:11:58.310 align:middle line:84% AUDIENCE: So did we agree that the "her" in this thing 00:11:58.310 --> 00:12:01.810 align:middle line:84% can refer to Mary potentially or are we saying specifically not? 00:12:01.810 --> 00:12:03.560 align:middle line:84% NORVIN RICHARDS: I asserted that it could. 00:12:03.560 --> 00:12:07.190 align:middle line:84% AUDIENCE: So it's that the NP contains "who knows her" 00:12:07.190 --> 00:12:08.990 align:middle line:90% and it's merging with the T bar. 00:12:08.990 --> 00:12:10.920 align:middle line:90% The T bar dominates "Mary." 00:12:10.920 --> 00:12:15.580 align:middle line:84% Then does that not contradict the new rule we just learned, 00:12:15.580 --> 00:12:17.783 align:middle line:84% or is it like the who knows part that combined with? 00:12:17.783 --> 00:12:18.700 align:middle line:90% NORVIN RICHARDS: Yeah. 00:12:18.700 --> 00:12:21.880 align:middle line:84% So here, let me draw "anyone who knows her" 00:12:21.880 --> 00:12:22.940 align:middle line:90% a little more carefully. 00:12:22.940 --> 00:12:26.920 align:middle line:90% So here's "anyone." 00:12:26.920 --> 00:12:28.780 align:middle line:84% Here's the relative clauses that I'm 00:12:28.780 --> 00:12:31.940 align:middle line:84% going to continue possibly foolishly not to label, 00:12:31.940 --> 00:12:33.340 align:middle line:90% so there's some stuff in here. 00:12:33.340 --> 00:12:36.970 align:middle line:84% And then, "who knows her," there's 00:12:36.970 --> 00:12:41.890 align:middle line:84% maybe a TP down in there containing a T bar and a T. 00:12:41.890 --> 00:12:44.240 align:middle line:84% And who knows, maybe the subject of it is "who." 00:12:44.240 --> 00:12:44.740 align:middle line:90% Right? 00:12:44.740 --> 00:12:46.660 align:middle line:84% So somewhere inside-- so there's "anyone" 00:12:46.660 --> 00:12:49.150 align:middle line:84% and then there's this relative clause which contains 00:12:49.150 --> 00:12:51.280 align:middle line:90% a sentence, "who knows her." 00:12:51.280 --> 00:12:53.080 align:middle line:84% That's what relative clauses are. 00:12:53.080 --> 00:12:54.760 align:middle line:84% They're a special kind of sentence 00:12:54.760 --> 00:12:57.310 align:middle line:84% that acts as a modifier for nouns. 00:12:57.310 --> 00:12:59.470 align:middle line:90% So you're right. 00:12:59.470 --> 00:13:05.050 align:middle line:84% And then this thing is becoming part of a larger TP. 00:13:05.050 --> 00:13:07.480 align:middle line:84% "Would"-- so now I'm just drawing 00:13:07.480 --> 00:13:12.250 align:middle line:84% what's on the slide with a little more detail-- 00:13:12.250 --> 00:13:16.360 align:middle line:90% "recognize"-- bless you-- 00:13:16.360 --> 00:13:18.760 align:middle line:90% "Mary." 00:13:18.760 --> 00:13:21.782 align:middle line:90% So now there are-- 00:13:21.782 --> 00:13:22.990 align:middle line:90% apart from "Mary"-- well, OK. 00:13:22.990 --> 00:13:24.907 align:middle line:84% There are three noun phrases in this sentence. 00:13:24.907 --> 00:13:28.060 align:middle line:84% There's "Mary," there's this thing, 00:13:28.060 --> 00:13:30.160 align:middle line:90% and there's this pronoun, "her." 00:13:30.160 --> 00:13:31.480 align:middle line:90% Yeah? 00:13:31.480 --> 00:13:37.450 align:middle line:84% And claim, a noun phrase can't refer 00:13:37.450 --> 00:13:41.440 align:middle line:84% to something that is contained in the alpha with which 00:13:41.440 --> 00:13:43.690 align:middle line:90% the noun phrase merges. 00:13:43.690 --> 00:13:47.110 align:middle line:84% So that's a claim that says when we ask, 00:13:47.110 --> 00:13:49.210 align:middle line:90% who does "her" refer to here? 00:13:49.210 --> 00:13:50.980 align:middle line:90% Can it refer to Mary? 00:13:50.980 --> 00:13:55.255 align:middle line:84% The answer is, yeah, because this was merged with "knows." 00:13:55.255 --> 00:13:58.960 align:middle line:84% "Knows" doesn't contain "Mary" and so we're all set. 00:13:58.960 --> 00:14:01.295 align:middle line:84% AUDIENCE: So your ancestors merging doesn't impact that? 00:14:01.295 --> 00:14:02.170 align:middle line:90% NORVIN RICHARDS: Yes. 00:14:02.170 --> 00:14:04.670 align:middle line:84% No, you're not responsible for the crimes of your ancestors. 00:14:04.670 --> 00:14:05.770 align:middle line:90% That's right. 00:14:05.770 --> 00:14:11.850 align:middle line:84% So this noun phrase shouldn't be able to refer to Mary. 00:14:11.850 --> 00:14:12.730 align:middle line:90% Yeah? 00:14:12.730 --> 00:14:14.240 align:middle line:90% And that might be true. 00:14:14.240 --> 00:14:15.980 align:middle line:90% Yeah. 00:14:15.980 --> 00:14:17.270 align:middle line:90% Yeah. 00:14:17.270 --> 00:14:18.630 align:middle line:90% Yeah? 00:14:18.630 --> 00:14:22.750 align:middle line:84% OK, so this has been an attempt at an argument form, argument 00:14:22.750 --> 00:14:24.860 align:middle line:90% by successful account. 00:14:24.860 --> 00:14:31.270 align:middle line:84% So I've shown you how, with this principle and these trees, 00:14:31.270 --> 00:14:33.340 align:middle line:84% the trees that I had on the slide, 00:14:33.340 --> 00:14:38.350 align:middle line:84% this kind of tree where a TP contains three things. 00:14:38.350 --> 00:14:41.710 align:middle line:84% There's a T, the sister of the T is the verb phrase, 00:14:41.710 --> 00:14:45.250 align:middle line:84% and then the daughter of the TP is 00:14:45.250 --> 00:14:46.542 align:middle line:90% the subject of the whole thing. 00:14:46.542 --> 00:14:48.500 align:middle line:84% That's the kind of tree that I was drawing you. 00:14:48.500 --> 00:14:50.740 align:middle line:84% Or to say that in another way, when you're assembling 00:14:50.740 --> 00:14:53.032 align:middle line:84% a sentence out of those three things, what you're doing 00:14:53.032 --> 00:14:55.150 align:middle line:84% is you're first merging T with the verb phrase 00:14:55.150 --> 00:14:57.520 align:middle line:84% and then you're merging the result of that with a noun 00:14:57.520 --> 00:14:58.300 align:middle line:90% phrase. 00:14:58.300 --> 00:15:00.790 align:middle line:84% That was the order in which I wanted to do things. 00:15:00.790 --> 00:15:03.430 align:middle line:84% I said, what I've tried to show you is that if we do things 00:15:03.430 --> 00:15:06.070 align:middle line:84% in that order we can have an account using 00:15:06.070 --> 00:15:09.430 align:middle line:84% this principle that I've just invoked 00:15:09.430 --> 00:15:11.752 align:middle line:84% for this contrast between these two sentences, the fact 00:15:11.752 --> 00:15:13.960 align:middle line:84% that "she" can't refer to Mary in the first sentence, 00:15:13.960 --> 00:15:18.250 align:middle line:84% but that "her" can refer to Mary in the second sentence. 00:15:18.250 --> 00:15:20.770 align:middle line:84% So this is, as I said, argument form, argument 00:15:20.770 --> 00:15:21.985 align:middle line:90% by successful account. 00:15:21.985 --> 00:15:24.310 align:middle line:84% If you buy these trees then, hey, 00:15:24.310 --> 00:15:25.870 align:middle line:84% we get this account of this contrast 00:15:25.870 --> 00:15:27.070 align:middle line:90% between these sentences. 00:15:27.070 --> 00:15:27.580 align:middle line:90% Yeah? 00:15:27.580 --> 00:15:33.010 align:middle line:84% Now suppose we were instead to do this kind of tree. 00:15:33.010 --> 00:15:35.770 align:middle line:90% Well, good. 00:15:35.770 --> 00:15:40.000 align:middle line:84% But notice at least the principle that I posited here 00:15:40.000 --> 00:15:43.100 align:middle line:84% wouldn't work for this kind of tree. 00:15:43.100 --> 00:15:44.830 align:middle line:84% So the principle that I've posited here 00:15:44.830 --> 00:15:48.800 align:middle line:84% says if a noun phrase is merged with some alpha the noun 00:15:48.800 --> 00:15:50.800 align:middle line:84% phrase can't refer to anything inside the alpha. 00:15:50.800 --> 00:15:53.950 align:middle line:84% Well, this noun phrase, "she" has just been merged with the T 00:15:53.950 --> 00:15:55.540 align:middle line:90% "would." 00:15:55.540 --> 00:15:59.920 align:middle line:84% So that principle at least would not draw this distinction. 00:15:59.920 --> 00:16:01.690 align:middle line:84% Possible response could be, yeah, but now 00:16:01.690 --> 00:16:04.060 align:middle line:90% we need another principle. 00:16:04.060 --> 00:16:08.140 align:middle line:84% To which the response is, yeah, well, what principle? 00:16:08.140 --> 00:16:09.570 align:middle line:90% Show it to me. 00:16:09.570 --> 00:16:12.010 align:middle line:84% So what we're going to do is we'll have this tree 00:16:12.010 --> 00:16:14.627 align:middle line:84% and we're comparing this tree with-- 00:16:14.627 --> 00:16:15.460 align:middle line:90% I'll draw it again-- 00:16:15.460 --> 00:16:18.640 align:middle line:90% 00:16:18.640 --> 00:16:25.720 align:middle line:84% "Mary," "recognize," "would," and now we're 00:16:25.720 --> 00:16:33.880 align:middle line:84% going to have "anyone," blah, blah, blah, "her." 00:16:33.880 --> 00:16:36.820 align:middle line:84% So we want this "her" to be able to refer to Mary 00:16:36.820 --> 00:16:39.895 align:middle line:84% but we want this "she" not to be able to refer to Mary. 00:16:39.895 --> 00:16:43.280 align:middle line:90% 00:16:43.280 --> 00:16:46.160 align:middle line:90% So there. 00:16:46.160 --> 00:16:48.530 align:middle line:84% That's part of an argument, anyway, 00:16:48.530 --> 00:16:51.380 align:middle line:84% that you should buy the kinds of trees, 00:16:51.380 --> 00:16:55.400 align:middle line:84% do the mergers in the order that I wanted you to do them in. 00:16:55.400 --> 00:16:58.100 align:middle line:84% We have a name for this relation that 00:16:58.100 --> 00:17:04.849 align:middle line:84% holds between "she" and "Mary" in this kind of tree. 00:17:04.849 --> 00:17:08.240 align:middle line:84% The relation that holds between a node like the she-- 00:17:08.240 --> 00:17:13.310 align:middle line:84% the noun phrase "she" in this tree and another node like 00:17:13.310 --> 00:17:17.510 align:middle line:84% the noun phrase "Mary" in this tree, which is contained-- 00:17:17.510 --> 00:17:18.400 align:middle line:90% I'm sorry. 00:17:18.400 --> 00:17:19.400 align:middle line:90% This is too complicated. 00:17:19.400 --> 00:17:20.895 align:middle line:90% Let me introduce some variables. 00:17:20.895 --> 00:17:22.520 align:middle line:84% We have a name for the kind of relation 00:17:22.520 --> 00:17:26.900 align:middle line:84% that holds between a node like the noun phrase "she"-- 00:17:26.900 --> 00:17:28.550 align:middle line:90% let's call that node x-- 00:17:28.550 --> 00:17:32.570 align:middle line:84% and another node that is contained in something 00:17:32.570 --> 00:17:34.910 align:middle line:90% that x was merged with. 00:17:34.910 --> 00:17:37.320 align:middle line:84% So the relation that holds between "she" and "Mary." 00:17:37.320 --> 00:17:37.820 align:middle line:90% Yeah. 00:17:37.820 --> 00:17:40.610 align:middle line:84% Or for that matter, between "she" and the verb phrase 00:17:40.610 --> 00:17:41.940 align:middle line:90% or between "she" and "would." 00:17:41.940 --> 00:17:42.440 align:middle line:90% Yeah? 00:17:42.440 --> 00:17:47.870 align:middle line:84% So the relationship that holds between a node x and everything 00:17:47.870 --> 00:17:52.230 align:middle line:84% that's inside the node with which that x was merged. 00:17:52.230 --> 00:17:52.730 align:middle line:90% Yeah. 00:17:52.730 --> 00:17:54.425 align:middle line:84% That relation is called C command. 00:17:54.425 --> 00:17:58.970 align:middle line:90% 00:17:58.970 --> 00:18:00.860 align:middle line:84% We're now not going to talk about C command 00:18:00.860 --> 00:18:03.020 align:middle line:84% for a little while because, as I said, 00:18:03.020 --> 00:18:05.060 align:middle line:84% you're experiencing a moment of anachronism. 00:18:05.060 --> 00:18:07.310 align:middle line:84% I'm telling you things that I was planning to tell you 00:18:07.310 --> 00:18:10.880 align:middle line:84% in a few weeks but because you demanded it, 00:18:10.880 --> 00:18:13.070 align:middle line:84% I'm offering you an argument for this particular way 00:18:13.070 --> 00:18:14.210 align:middle line:90% of assembling trees. 00:18:14.210 --> 00:18:15.442 align:middle line:90% Yeah. 00:18:15.442 --> 00:18:17.390 align:middle line:90% OK? 00:18:17.390 --> 00:18:19.298 align:middle line:90% All right. 00:18:19.298 --> 00:18:19.965 align:middle line:90% All right, good. 00:18:19.965 --> 00:18:23.850 align:middle line:90% 00:18:23.850 --> 00:18:24.360 align:middle line:90% Yes. 00:18:24.360 --> 00:18:25.500 align:middle line:84% Alternative way of drawing trees wouldn't 00:18:25.500 --> 00:18:26.820 align:middle line:90% let us use this explanation. 00:18:26.820 --> 00:18:28.830 align:middle line:84% You would need a different one, and then 00:18:28.830 --> 00:18:31.233 align:middle line:84% if you like that alternative way of drawing trees, 00:18:31.233 --> 00:18:33.150 align:middle line:84% it's your job to come up with a different one. 00:18:33.150 --> 00:18:36.275 align:middle line:84% Kateryna, yes, has already come up with one. 00:18:36.275 --> 00:18:37.650 align:middle line:84% AUDIENCE: I just want to clarify. 00:18:37.650 --> 00:18:39.585 align:middle line:84% Does C command refer to that principle 00:18:39.585 --> 00:18:40.710 align:middle line:90% that we were talking about? 00:18:40.710 --> 00:18:42.002 align:middle line:90% NORVIN RICHARDS: No, I'm sorry. 00:18:42.002 --> 00:18:44.670 align:middle line:90% Let me say it better. 00:18:44.670 --> 00:18:47.220 align:middle line:84% There's a definition for C command. 00:18:47.220 --> 00:18:47.300 align:middle line:84% AUDIENCE: C command is the relationship? 00:18:47.300 --> 00:18:49.258 align:middle line:84% NORVIN RICHARDS: C command is the relationship. 00:18:49.258 --> 00:18:55.350 align:middle line:90% So you say "x C commands y." 00:18:55.350 --> 00:18:57.960 align:middle line:84% There are various ways to define C command, but one of them-- 00:18:57.960 --> 00:19:00.585 align:middle line:84% here, I'll show you one and then we can talk about another one. 00:19:00.585 --> 00:19:06.900 align:middle line:84% You can say "x C commands y if every node that dominates 00:19:06.900 --> 00:19:16.670 align:middle line:84% x dominates y," and people usually add 00:19:16.670 --> 00:19:19.430 align:middle line:90% "and x does not dominate y." 00:19:19.430 --> 00:19:23.750 align:middle line:90% 00:19:23.750 --> 00:19:25.730 align:middle line:84% So this is one way to define it formally. 00:19:25.730 --> 00:19:26.600 align:middle line:90% What does that do? 00:19:26.600 --> 00:19:28.850 align:middle line:84% It says if you look at the noun phrase 00:19:28.850 --> 00:19:34.070 align:middle line:84% "she," what the noun phrase C commands is everything that 00:19:34.070 --> 00:19:37.130 align:middle line:84% is dominated by everything that dominates "she," 00:19:37.130 --> 00:19:38.330 align:middle line:90% the noun phrase "she." 00:19:38.330 --> 00:19:40.280 align:middle line:84% So what dominates the noun phrase "she"? 00:19:40.280 --> 00:19:43.430 align:middle line:90% Well, it's just the node TP. 00:19:43.430 --> 00:19:45.890 align:middle line:84% And so "she" C commands everything that's 00:19:45.890 --> 00:19:48.890 align:middle line:84% dominated by the node TP, which is basically 00:19:48.890 --> 00:19:50.180 align:middle line:90% everything except for TP. 00:19:50.180 --> 00:19:53.150 align:middle line:84% So T bar and T and the verb phrase. 00:19:53.150 --> 00:19:57.380 align:middle line:84% The T would C commands the verb phrase 00:19:57.380 --> 00:20:00.110 align:middle line:84% and also the verb and the noun phrase "Mary." 00:20:00.110 --> 00:20:01.070 align:middle line:90% Yeah. 00:20:01.070 --> 00:20:03.020 align:middle line:84% To put it yet another way, you C command 00:20:03.020 --> 00:20:06.350 align:middle line:84% your sister and everything your sister dominates. 00:20:06.350 --> 00:20:08.810 align:middle line:90% That's what C command is. 00:20:08.810 --> 00:20:11.450 align:middle line:84% Lots and lots of syntax makes reference to C command, 00:20:11.450 --> 00:20:13.530 align:middle line:84% and so we will keep coming back to it. 00:20:13.530 --> 00:20:16.860 align:middle line:84% It's a thing that will come up again. 00:20:16.860 --> 00:20:18.120 align:middle line:90% OK? 00:20:18.120 --> 00:20:21.570 align:middle line:84% So again, this is all dramatic foreshadowing or anachronism 00:20:21.570 --> 00:20:22.920 align:middle line:90% or whatever you want to say-- 00:20:22.920 --> 00:20:24.390 align:middle line:90% whatever you want to call it. 00:20:24.390 --> 00:20:26.280 align:middle line:84% These are things I was planning to teach you 00:20:26.280 --> 00:20:29.340 align:middle line:84% a little bit in the future, and so now you 00:20:29.340 --> 00:20:31.830 align:middle line:90% have a jump on in life. 00:20:31.830 --> 00:20:34.860 align:middle line:84% Later on some things will suddenly become familiar. 00:20:34.860 --> 00:20:35.610 align:middle line:90% Yeah. 00:20:35.610 --> 00:20:38.260 align:middle line:90% Yep. 00:20:38.260 --> 00:20:38.950 align:middle line:90% All right, good. 00:20:38.950 --> 00:20:41.840 align:middle line:90% 00:20:41.840 --> 00:20:44.378 align:middle line:84% I have been teasing you for making me do this 00:20:44.378 --> 00:20:45.920 align:middle line:84% but I really appreciate it, actually. 00:20:45.920 --> 00:20:48.620 align:middle line:84% It's a real pleasure teaching people who 00:20:48.620 --> 00:20:50.375 align:middle line:90% disagree with me all the time. 00:20:50.375 --> 00:20:58.743 align:middle line:90% 00:20:58.743 --> 00:21:01.160 align:middle line:84% When I'm on airplanes and I can't avoid telling the person 00:21:01.160 --> 00:21:02.840 align:middle line:84% next to me what I do for a living and I tell them 00:21:02.840 --> 00:21:04.490 align:middle line:84% I teach linguistics at MIT, they're like, 00:21:04.490 --> 00:21:06.320 align:middle line:84% oh-- sometimes they say, "Oh, I didn't know 00:21:06.320 --> 00:21:07.528 align:middle line:90% they had linguistics at MIT." 00:21:07.528 --> 00:21:09.380 align:middle line:90% I'm like, "Yes, yes!" 00:21:09.380 --> 00:21:13.430 align:middle line:84% And then they say, "So what's it like teaching MIT undergrads 00:21:13.430 --> 00:21:14.090 align:middle line:90% linguistics?" 00:21:14.090 --> 00:21:15.715 align:middle line:84% And I say, "It's great because they all 00:21:15.715 --> 00:21:17.690 align:middle line:84% have these math and science backgrounds so they 00:21:17.690 --> 00:21:19.010 align:middle line:90% don't believe a word I say. 00:21:19.010 --> 00:21:20.570 align:middle line:90% It's terrific. 00:21:20.570 --> 00:21:22.580 align:middle line:84% So I go in there and I tell them things 00:21:22.580 --> 00:21:25.210 align:middle line:84% and they're like, well, no, show me some proof. 00:21:25.210 --> 00:21:27.578 align:middle line:90% I go, well, OK. 00:21:27.578 --> 00:21:29.120 align:middle line:84% In another university I think I would 00:21:29.120 --> 00:21:31.520 align:middle line:84% be able to get away with just the authority of having 00:21:31.520 --> 00:21:33.260 align:middle line:90% a beard this length. 00:21:33.260 --> 00:21:36.770 align:middle line:84% People would just believe me, but no. 00:21:36.770 --> 00:21:37.730 align:middle line:90% It's great." 00:21:37.730 --> 00:21:39.020 align:middle line:90% OK. 00:21:39.020 --> 00:21:40.530 align:middle line:90% This is a slide we saw before. 00:21:40.530 --> 00:21:43.440 align:middle line:84% This is just to remind you of a bunch of terminology. 00:21:43.440 --> 00:21:45.830 align:middle line:84% Some of it I've been using as we've been talking. 00:21:45.830 --> 00:21:50.960 align:middle line:84% So just to remind you, when we're talking about trees, 00:21:50.960 --> 00:21:52.670 align:middle line:84% standardly use feminine kinship terms 00:21:52.670 --> 00:21:54.770 align:middle line:84% to talk about relations between parts of trees. 00:21:54.770 --> 00:21:58.790 align:middle line:84% So we talk about sisters, which are two things that are both-- 00:21:58.790 --> 00:22:01.370 align:middle line:84% two things that were merged to form a node. 00:22:01.370 --> 00:22:04.580 align:middle line:84% So for example, the D, "the," and the N, 00:22:04.580 --> 00:22:06.560 align:middle line:90% "child," those are sisters. 00:22:06.560 --> 00:22:08.540 align:middle line:84% The V, "tickle," and the noun phrase, 00:22:08.540 --> 00:22:12.350 align:middle line:84% "the child," those are sisters in this tree. 00:22:12.350 --> 00:22:13.400 align:middle line:90% We talk about mothers. 00:22:13.400 --> 00:22:17.660 align:middle line:84% So the noun phrase "the child" is the mother of D and N, 00:22:17.660 --> 00:22:19.860 align:middle line:90% "the" and "child." 00:22:19.860 --> 00:22:21.600 align:middle line:90% We also talk about daughters. 00:22:21.600 --> 00:22:25.350 align:middle line:84% So "the" and "child" are the daughters of the noun phrase. 00:22:25.350 --> 00:22:27.077 align:middle line:84% And I think I said last time that's it, 00:22:27.077 --> 00:22:28.910 align:middle line:84% so we don't talk about grandmothers or aunts 00:22:28.910 --> 00:22:30.690 align:middle line:90% or anything else. 00:22:30.690 --> 00:22:34.640 align:middle line:84% And then and this came up when Giday asked a question earlier. 00:22:34.640 --> 00:22:38.390 align:middle line:84% The basic relation that's reflected in these trees 00:22:38.390 --> 00:22:41.060 align:middle line:84% is the relation of immediate domination. 00:22:41.060 --> 00:22:43.260 align:middle line:84% That's the relation that's created by the operation 00:22:43.260 --> 00:22:43.760 align:middle line:90% "merge." 00:22:43.760 --> 00:22:47.060 align:middle line:84% So when you merge two things, you 00:22:47.060 --> 00:22:49.580 align:middle line:84% create a new thing that immediately dominates the two 00:22:49.580 --> 00:22:50.570 align:middle line:90% things that you merged. 00:22:50.570 --> 00:22:54.230 align:middle line:84% That's just our name for that relation in the tree. 00:22:54.230 --> 00:22:57.830 align:middle line:84% And then there is another notion, dominate, 00:22:57.830 --> 00:23:00.780 align:middle line:84% which is the transitive closure of immediately dominate. 00:23:00.780 --> 00:23:03.353 align:middle line:84% So you dominate the things that you immediately dominate 00:23:03.353 --> 00:23:06.020 align:middle line:84% and the things that those things immediately dominate and so on, 00:23:06.020 --> 00:23:08.580 align:middle line:90% all the way down to the tree. 00:23:08.580 --> 00:23:11.670 align:middle line:84% And then we said there's this word "constituent," which we've 00:23:11.670 --> 00:23:15.552 align:middle line:84% been using for some time, and all of you 00:23:15.552 --> 00:23:17.010 align:middle line:84% seem to have a handle on it anyway. 00:23:17.010 --> 00:23:19.080 align:middle line:84% But just to give it a formal definition, 00:23:19.080 --> 00:23:21.060 align:middle line:84% we'll say that something is a constituent 00:23:21.060 --> 00:23:24.060 align:middle line:84% if there's a single node that dominates 00:23:24.060 --> 00:23:26.890 align:middle line:90% all the words in alpha. 00:23:26.890 --> 00:23:30.060 align:middle line:84% So "the child" is a constituent because there's 00:23:30.060 --> 00:23:33.660 align:middle line:84% a node that dominates just those words, the noun phrase. 00:23:33.660 --> 00:23:36.030 align:middle line:84% "Tickle the child" is a constituent 00:23:36.030 --> 00:23:39.930 align:middle line:84% because there's a verb phrase dominating just those words. 00:23:39.930 --> 00:23:42.480 align:middle line:84% "Will tickle the" is not a constituent 00:23:42.480 --> 00:23:46.220 align:middle line:84% because there's no node dominating just those words. 00:23:46.220 --> 00:23:48.190 align:middle line:90% Is that clear? 00:23:48.190 --> 00:23:51.150 align:middle line:84% This is all just terminology, and I'm 00:23:51.150 --> 00:23:53.650 align:middle line:84% telling you about it because you're going to hear me use it. 00:23:53.650 --> 00:23:56.150 align:middle line:90% In fact, you already have. 00:23:56.150 --> 00:23:56.870 align:middle line:90% OK. 00:23:56.870 --> 00:23:57.650 align:middle line:90% All right? 00:23:57.650 --> 00:24:02.670 align:middle line:84% So the way we've been talking, then, we take pairs of things, 00:24:02.670 --> 00:24:05.760 align:middle line:84% we merge them to create new things. 00:24:05.760 --> 00:24:08.320 align:middle line:90% And we haven't said very much. 00:24:08.320 --> 00:24:10.200 align:middle line:84% In fact, we haven't said anything at all 00:24:10.200 --> 00:24:14.430 align:middle line:84% about why you pick the particular things 00:24:14.430 --> 00:24:16.150 align:middle line:90% that you pick. 00:24:16.150 --> 00:24:18.660 align:middle line:84% So I want to talk about that a little bit. 00:24:18.660 --> 00:24:23.430 align:middle line:84% Here are some sentences mostly involving ants, I guess, 00:24:23.430 --> 00:24:27.600 align:middle line:84% and anteaters, And they're all grammatical sentences 00:24:27.600 --> 00:24:29.190 align:middle line:90% of English, I think. 00:24:29.190 --> 00:24:31.590 align:middle line:84% "The ants thrived" and "The anteater arrived"-- 00:24:31.590 --> 00:24:34.350 align:middle line:84% this is kind of a little dramatic story 00:24:34.350 --> 00:24:37.338 align:middle line:84% about some ants and their difficult lives!-- 00:24:37.338 --> 00:24:38.130 align:middle line:90% "The ants thrived." 00:24:38.130 --> 00:24:39.030 align:middle line:90% "The anteater arrived." 00:24:39.030 --> 00:24:40.440 align:middle line:84% "The anteater devoured the ants." 00:24:40.440 --> 00:24:41.773 align:middle line:90% And "Mary slapped the anteater." 00:24:41.773 --> 00:24:43.920 align:middle line:84% A little sad story about ants, or a happy story 00:24:43.920 --> 00:24:44.670 align:middle line:90% about an anteater. 00:24:44.670 --> 00:24:46.780 align:middle line:84% It depends on whose side you're on, I guess. 00:24:46.780 --> 00:24:47.280 align:middle line:90% Yeah? 00:24:47.280 --> 00:24:50.920 align:middle line:90% 00:24:50.920 --> 00:24:53.530 align:middle line:90% Observation, old observation. 00:24:53.530 --> 00:24:59.210 align:middle line:84% Verbs, in lots of languages, maybe all languages, 00:24:59.210 --> 00:25:03.270 align:middle line:84% can be fruitfully thought of as coming in at least two types. 00:25:03.270 --> 00:25:06.410 align:middle line:84% There are verbs like "thrive" and "arrive" 00:25:06.410 --> 00:25:09.860 align:middle line:90% which need to not have objects. 00:25:09.860 --> 00:25:12.110 align:middle line:84% So you can say "The ants thrived" or "The anteater 00:25:12.110 --> 00:25:12.830 align:middle line:90% arrived." 00:25:12.830 --> 00:25:15.170 align:middle line:84% You cannot say "The ants thrived the ant farm" 00:25:15.170 --> 00:25:17.330 align:middle line:84% or "The anteater arrived the anthill." 00:25:17.330 --> 00:25:21.220 align:middle line:84% You can't put an object after those kinds of verbs. 00:25:21.220 --> 00:25:24.430 align:middle line:84% And then there are other verbs, like "devour" and "slap," 00:25:24.430 --> 00:25:25.810 align:middle line:90% that are the opposite. 00:25:25.810 --> 00:25:27.730 align:middle line:84% You can say "The anteater devoured the ants." 00:25:27.730 --> 00:25:29.730 align:middle line:84% You can't very well say "The anteater devoured." 00:25:29.730 --> 00:25:31.430 align:middle line:90% That sentence is not finished. 00:25:31.430 --> 00:25:32.140 align:middle line:90% Yeah. 00:25:32.140 --> 00:25:34.360 align:middle line:84% So really "Mary slapped the anteater." 00:25:34.360 --> 00:25:36.730 align:middle line:90% You can't say "Mary slapped." 00:25:36.730 --> 00:25:39.070 align:middle line:84% Did I hear noises of disagreement? 00:25:39.070 --> 00:25:41.412 align:middle line:84% Facial expressions of disagreement? 00:25:41.412 --> 00:25:42.680 align:middle line:90% AUDIENCE: "Mary slaps." 00:25:42.680 --> 00:25:45.080 align:middle line:90% NORVIN RICHARDS: Sorry? 00:25:45.080 --> 00:25:48.260 align:middle line:84% Apart from the technical term, yeah. 00:25:48.260 --> 00:25:49.550 align:middle line:90% Apart from that. 00:25:49.550 --> 00:25:51.760 align:middle line:90% Yeah? 00:25:51.760 --> 00:25:53.020 align:middle line:90% OK. 00:25:53.020 --> 00:25:55.793 align:middle line:84% So there are verbs that seem to want to have objects 00:25:55.793 --> 00:25:58.210 align:middle line:84% and there are verbs that seem to want to not have objects. 00:25:58.210 --> 00:25:59.230 align:middle line:90% Classic observation. 00:25:59.230 --> 00:26:00.790 align:middle line:84% The name for that classic observation 00:26:00.790 --> 00:26:03.520 align:middle line:84% is that there are verbs that are what's called transitive. 00:26:03.520 --> 00:26:05.440 align:middle line:84% Those are the verbs that want to have objects, 00:26:05.440 --> 00:26:06.940 align:middle line:90% verbs like "devour" and "slap." 00:26:06.940 --> 00:26:08.773 align:middle line:84% There are other verbs that are intransitive. 00:26:08.773 --> 00:26:11.350 align:middle line:84% Those are verbs like "thrive" and "arrive." 00:26:11.350 --> 00:26:13.450 align:middle line:84% I picked these verbs fairly carefully. 00:26:13.450 --> 00:26:15.790 align:middle line:90% There are many, many-- 00:26:15.790 --> 00:26:19.180 align:middle line:84% possibly most verbs-- that can be either one. 00:26:19.180 --> 00:26:23.320 align:middle line:84% So for example, you can say "The anteater ate the ants" 00:26:23.320 --> 00:26:25.660 align:middle line:84% and you can also say "The anteater ate." 00:26:25.660 --> 00:26:26.440 align:middle line:90% It's fine. 00:26:26.440 --> 00:26:28.660 align:middle line:90% Yeah? 00:26:28.660 --> 00:26:32.350 align:middle line:84% So some verbs are transitive, others are intransitive. 00:26:32.350 --> 00:26:35.330 align:middle line:84% Maybe there are also other verbs that are indecisive. 00:26:35.330 --> 00:26:35.830 align:middle line:90% Yeah? 00:26:35.830 --> 00:26:37.240 align:middle line:84% AUDIENCE: In that context, how is "ate" 00:26:37.240 --> 00:26:38.350 align:middle line:90% different from "devoured"? 00:26:38.350 --> 00:26:39.100 align:middle line:90% NORVIN RICHARDS: Good question. 00:26:39.100 --> 00:26:39.600 align:middle line:90% Yeah. 00:26:39.600 --> 00:26:42.470 align:middle line:84% But I think the fact, the facts are as I just described them, 00:26:42.470 --> 00:26:42.970 align:middle line:90% right? 00:26:42.970 --> 00:26:45.920 align:middle line:84% That "The anteater devoured" is not a complete sentence, 00:26:45.920 --> 00:26:47.830 align:middle line:90% but "The anteater ate" is. 00:26:47.830 --> 00:26:50.620 align:middle line:84% "Devoured" and "ate" mean slightly different things. 00:26:50.620 --> 00:26:54.665 align:middle line:84% You have to be louder and messier and more violent 00:26:54.665 --> 00:26:56.290 align:middle line:84% if you're devouring, I guess, and "eat" 00:26:56.290 --> 00:26:57.912 align:middle line:90% is a more plain vanilla verb. 00:26:57.912 --> 00:27:00.370 align:middle line:84% Why that means that one of these verbs has to be transitive 00:27:00.370 --> 00:27:02.370 align:middle line:84% and the other doesn't is a really good question. 00:27:02.370 --> 00:27:03.710 align:middle line:90% Yeah. 00:27:03.710 --> 00:27:04.880 align:middle line:90% Yeah. 00:27:04.880 --> 00:27:07.380 align:middle line:90% Other questions? 00:27:07.380 --> 00:27:09.660 align:middle line:84% To which I may, again, only be able to say, yeah, 00:27:09.660 --> 00:27:11.230 align:middle line:90% that's a good question. 00:27:11.230 --> 00:27:11.730 align:middle line:90% Yeah? 00:27:11.730 --> 00:27:14.238 align:middle line:84% AUDIENCE: I was just going to follow up and then say, 00:27:14.238 --> 00:27:15.780 align:middle line:84% are we going to answer that question? 00:27:15.780 --> 00:27:16.613 align:middle line:90% NORVIN RICHARDS: No. 00:27:16.613 --> 00:27:17.400 align:middle line:90% No, no, no. 00:27:17.400 --> 00:27:18.180 align:middle line:90% No. 00:27:18.180 --> 00:27:19.140 align:middle line:90% No. 00:27:19.140 --> 00:27:20.400 align:middle line:90% No, we are not. 00:27:20.400 --> 00:27:21.720 align:middle line:90% Other questions? 00:27:21.720 --> 00:27:23.310 align:middle line:84% I enjoy listening to your questions 00:27:23.310 --> 00:27:26.820 align:middle line:90% even if I can't answer them. 00:27:26.820 --> 00:27:28.320 align:middle line:84% OK, so some verbs are transitive, 00:27:28.320 --> 00:27:30.870 align:middle line:90% others are intransitive. 00:27:30.870 --> 00:27:32.475 align:middle line:90% So that's a classic observation. 00:27:32.475 --> 00:27:34.350 align:middle line:84% People have been saying that since the Romans 00:27:34.350 --> 00:27:36.210 align:middle line:90% and the Greeks. 00:27:36.210 --> 00:27:38.190 align:middle line:84% Less classic observation is that there 00:27:38.190 --> 00:27:41.820 align:middle line:84% are other kinds of verbs that not only need an object 00:27:41.820 --> 00:27:43.710 align:middle line:84% but also need a prepositional phrase. 00:27:43.710 --> 00:27:44.820 align:middle line:84% Actually, there are verbs that just 00:27:44.820 --> 00:27:45.987 align:middle line:90% need a prepositional phrase. 00:27:45.987 --> 00:27:47.800 align:middle line:84% We'll talk about that maybe later, 00:27:47.800 --> 00:27:50.856 align:middle line:84% but let's talk for now about a verb like "put." 00:27:50.856 --> 00:27:54.280 align:middle line:84% "Put" absolutely has to have an object, 00:27:54.280 --> 00:27:55.710 align:middle line:84% so it's transitive in that sense. 00:27:55.710 --> 00:27:58.630 align:middle line:84% But it also has to have a prepositional phrase. 00:27:58.630 --> 00:28:02.110 align:middle line:84% So you can say "The anteater put the ants onto a plate." 00:28:02.110 --> 00:28:05.490 align:middle line:84% This is a comparatively civilized anteater. 00:28:05.490 --> 00:28:07.923 align:middle line:84% But you can't leave out either the object 00:28:07.923 --> 00:28:09.090 align:middle line:90% or the prepositional phrase. 00:28:09.090 --> 00:28:11.340 align:middle line:84% You can't say "The anteater put the ants." 00:28:11.340 --> 00:28:12.390 align:middle line:90% That's not a sentence. 00:28:12.390 --> 00:28:15.870 align:middle line:84% And you also cannot say "The anteater put onto a plate." 00:28:15.870 --> 00:28:17.220 align:middle line:90% Also not a sentence. 00:28:17.220 --> 00:28:20.970 align:middle line:84% So "put" needs to combine with both an object 00:28:20.970 --> 00:28:23.850 align:middle line:90% and a preposition. 00:28:23.850 --> 00:28:26.217 align:middle line:90% So apparently, in-- 00:28:26.217 --> 00:28:28.050 align:middle line:84% I hope this is what I say on the next slide. 00:28:28.050 --> 00:28:30.523 align:middle line:90% Yeah. 00:28:30.523 --> 00:28:32.190 align:middle line:84% Remember when we were talking about what 00:28:32.190 --> 00:28:36.510 align:middle line:84% we would need to state in our lexicon about each word 00:28:36.510 --> 00:28:38.910 align:middle line:84% or about each morpheme, if we're putting morphemes 00:28:38.910 --> 00:28:42.795 align:middle line:84% in our lexicon, which we were saying probably we should? 00:28:42.795 --> 00:28:47.263 align:middle line:84% We were saying, yeah, When you list an entry in your lexicon 00:28:47.263 --> 00:28:48.930 align:middle line:84% you're going to need to say whether it's 00:28:48.930 --> 00:28:50.790 align:middle line:84% a free morpheme or a bound morpheme 00:28:50.790 --> 00:28:53.130 align:middle line:84% and whether it's a prefix or a suffix or an infix 00:28:53.130 --> 00:28:56.460 align:middle line:84% or whatever all else, and how it's pronounced. 00:28:56.460 --> 00:28:59.520 align:middle line:84% And we spend some time doing phonology and convincing 00:28:59.520 --> 00:29:01.830 align:middle line:84% ourselves that saying how something is pronounced 00:29:01.830 --> 00:29:04.320 align:middle line:84% might be more complicated than it looks. 00:29:04.320 --> 00:29:07.530 align:middle line:84% Here's something else that we're going to have to state. 00:29:07.530 --> 00:29:13.200 align:middle line:84% When we list a verb we're going to need to state something 00:29:13.200 --> 00:29:14.680 align:middle line:84% about-- this is called selection. 00:29:14.680 --> 00:29:17.410 align:middle line:84% So we need to say verbs, for example, 00:29:17.410 --> 00:29:21.180 align:middle line:84% select for the kinds of things that they want to combine with. 00:29:21.180 --> 00:29:25.452 align:middle line:84% So our list, our lexical entry for a verb like "devour" 00:29:25.452 --> 00:29:26.910 align:middle line:84% has to say, oh yeah, this is a verb 00:29:26.910 --> 00:29:28.118 align:middle line:90% that needs to have an object. 00:29:28.118 --> 00:29:30.500 align:middle line:90% 00:29:30.500 --> 00:29:31.730 align:middle line:90% I think we need to say that. 00:29:31.730 --> 00:29:34.460 align:middle line:84% This gets back to your question of a second ago, right? 00:29:34.460 --> 00:29:36.620 align:middle line:84% Can we derive it from anything else 00:29:36.620 --> 00:29:39.350 align:middle line:84% about the meaning of "devour" that it has to have an object? 00:29:39.350 --> 00:29:41.540 align:middle line:84% Can we make that follow from something? 00:29:41.540 --> 00:29:44.030 align:middle line:84% Rather than just having to say next to "devour," oh yeah, 00:29:44.030 --> 00:29:45.500 align:middle line:90% this one needs an object. 00:29:45.500 --> 00:29:48.000 align:middle line:84% Can that be a general property of certain classes of verbs? 00:29:48.000 --> 00:29:48.950 align:middle line:90% That's an ongoing project. 00:29:48.950 --> 00:29:51.117 align:middle line:84% There are linguists trying to figure out the answers 00:29:51.117 --> 00:29:53.460 align:middle line:90% to questions like that. 00:29:53.460 --> 00:29:56.740 align:middle line:84% But yeah, for 24.900 we'll just state for every verb, 00:29:56.740 --> 00:29:58.500 align:middle line:90% this verb needs an object. 00:29:58.500 --> 00:29:59.940 align:middle line:90% This is kind of like-- 00:29:59.940 --> 00:30:02.250 align:middle line:84% and it should look a little familiar-- 00:30:02.250 --> 00:30:07.830 align:middle line:84% when we were doing "unlockable" and other polymorphemic words 00:30:07.830 --> 00:30:10.650 align:middle line:84% like that we were saying, yeah, there are affixes that say, 00:30:10.650 --> 00:30:13.800 align:middle line:84% "I want to combine with a verb," let's say, 00:30:13.800 --> 00:30:17.680 align:middle line:84% and "I'll give you an adjective as a result." 00:30:17.680 --> 00:30:20.490 align:middle line:84% Similarly, verbs have to specify what they merge with. 00:30:20.490 --> 00:30:24.330 align:middle line:84% So the verb "put" is pronounced "put"-- 00:30:24.330 --> 00:30:26.550 align:middle line:84% I should really have done that in IPA. 00:30:26.550 --> 00:30:28.440 align:middle line:90% It's pronounced whatever. 00:30:28.440 --> 00:30:29.820 align:middle line:90% "Put." 00:30:29.820 --> 00:30:32.590 align:middle line:90% And it means something. 00:30:32.590 --> 00:30:34.560 align:middle line:84% It means to cause something to be in a place, 00:30:34.560 --> 00:30:36.685 align:middle line:84% whatever, however we're going to state its meaning, 00:30:36.685 --> 00:30:39.250 align:middle line:84% and it selects for a noun phrase and a prepositional phrase. 00:30:39.250 --> 00:30:40.500 align:middle line:90% That's the way we'll say that. 00:30:40.500 --> 00:30:44.150 align:middle line:84% So the verb put needs to merge with those things. 00:30:44.150 --> 00:30:44.650 align:middle line:90% OK? 00:30:44.650 --> 00:30:48.060 align:middle line:90% 00:30:48.060 --> 00:30:49.830 align:middle line:84% Actually, be a little more specific 00:30:49.830 --> 00:30:52.530 align:middle line:84% about how selection works exactly. 00:30:52.530 --> 00:30:56.628 align:middle line:84% So I just said-- yeah, is this much clearer so far? 00:30:56.628 --> 00:30:58.920 align:middle line:84% So let's be a little more specific about what selection 00:30:58.920 --> 00:31:03.000 align:middle line:84% is and what exactly you're selecting for. 00:31:03.000 --> 00:31:05.520 align:middle line:84% I just said there are verbs like "put" that 00:31:05.520 --> 00:31:08.040 align:middle line:84% select for an object and also a prepositional phrase. 00:31:08.040 --> 00:31:09.810 align:middle line:84% There are also verbs that only select 00:31:09.810 --> 00:31:11.100 align:middle line:90% for a prepositional phrase. 00:31:11.100 --> 00:31:13.380 align:middle line:90% "Depend" is one of those. 00:31:13.380 --> 00:31:18.130 align:middle line:90% So "depend" can combine-- 00:31:18.130 --> 00:31:24.250 align:middle line:84% kind of needs to combine, it depends, but at least can 00:31:24.250 --> 00:31:26.590 align:middle line:84% combine with a prepositional phrase, 00:31:26.590 --> 00:31:28.397 align:middle line:84% but not just any prepositional phrase. 00:31:28.397 --> 00:31:30.730 align:middle line:84% It needs to combine with a prepositional phrase in which 00:31:30.730 --> 00:31:33.460 align:middle line:90% the preposition is "on." 00:31:33.460 --> 00:31:35.800 align:middle line:90% You can depend on things. 00:31:35.800 --> 00:31:39.190 align:middle line:84% You cannot depend from things or at things or near things 00:31:39.190 --> 00:31:40.150 align:middle line:90% or by things. 00:31:40.150 --> 00:31:41.740 align:middle line:90% You can't depend near something. 00:31:41.740 --> 00:31:45.310 align:middle line:84% That's not how you use the word "depend." 00:31:45.310 --> 00:31:47.770 align:middle line:84% If you learn other languages-- especially 00:31:47.770 --> 00:31:49.570 align:middle line:84% if you learn other Indo-European languages 00:31:49.570 --> 00:31:51.070 align:middle line:84% because Indo-European languages seem 00:31:51.070 --> 00:31:56.580 align:middle line:84% to be especially fond of prepositions, I don't know, 00:31:56.580 --> 00:31:59.570 align:middle line:84% Indo-European languages, the language family that English is 00:31:59.570 --> 00:32:03.470 align:middle line:84% part of, we have more prepositions than languages 00:32:03.470 --> 00:32:04.290 align:middle line:90% really should-- 00:32:04.290 --> 00:32:06.830 align:middle line:84% And if you learn a new language, there's 00:32:06.830 --> 00:32:09.350 align:middle line:84% this weird mapping from one language to the next 00:32:09.350 --> 00:32:11.480 align:middle line:84% where you just need to learn which prepositions. 00:32:11.480 --> 00:32:12.410 align:middle line:90% It's not simple. 00:32:12.410 --> 00:32:14.270 align:middle line:84% You don't get to learn "This preposition is the translation 00:32:14.270 --> 00:32:15.260 align:middle line:90% for that preposition." 00:32:15.260 --> 00:32:17.480 align:middle line:84% You learn things like, "This is the preposition that 00:32:17.480 --> 00:32:21.680 align:middle line:84% means 'at' for cities and large islands but for small islands 00:32:21.680 --> 00:32:23.930 align:middle line:84% you should use this other preposition instead." 00:32:23.930 --> 00:32:25.280 align:middle line:84% Not to scare you or anything, but if you 00:32:25.280 --> 00:32:27.238 align:middle line:84% want to learn other Indo-European languages you 00:32:27.238 --> 00:32:29.632 align:middle line:84% should brace yourself for things like that. 00:32:29.632 --> 00:32:31.340 align:middle line:84% If you don't want to do things like that, 00:32:31.340 --> 00:32:32.900 align:middle line:84% learn a language that's not Indo-European, 00:32:32.900 --> 00:32:35.030 align:middle line:84% because outside Indo-European it's mostly like, 00:32:35.030 --> 00:32:36.470 align:middle line:84% "Here is our locative preposition. 00:32:36.470 --> 00:32:39.362 align:middle line:90% Please use responsibly." 00:32:39.362 --> 00:32:41.070 align:middle line:84% But English is an Indo-European language, 00:32:41.070 --> 00:32:43.640 align:middle line:84% and so we have stuff like this, verbs like "depend" that say, 00:32:43.640 --> 00:32:46.490 align:middle line:84% "I want a prepositional phrase and I want the preposition 00:32:46.490 --> 00:32:48.190 align:middle line:90% to be 'on.' 00:32:48.190 --> 00:32:49.190 align:middle line:90% Can't be anything else." 00:32:49.190 --> 00:32:52.280 align:middle line:90% 00:32:52.280 --> 00:32:55.360 align:middle line:84% So apparently it's possible, at least in some kinds of cases, 00:32:55.360 --> 00:32:58.670 align:middle line:84% for a verb to select not just a prepositional phrase, 00:32:58.670 --> 00:33:00.880 align:middle line:84% but a prepositional phrase with a particular head. 00:33:00.880 --> 00:33:03.370 align:middle line:84% Where by head I mean the preposition-- the thing 00:33:03.370 --> 00:33:05.995 align:middle line:84% that the prepositional phrase is named after, 00:33:05.995 --> 00:33:12.630 align:middle line:84% the smallest thing with the label P. 00:33:12.630 --> 00:33:15.090 align:middle line:84% So that's the kind of selection you get. 00:33:15.090 --> 00:33:17.910 align:middle line:84% There are other imaginable kinds of selection 00:33:17.910 --> 00:33:19.620 align:middle line:90% that you don't ever get. 00:33:19.620 --> 00:33:23.895 align:middle line:84% So prepositional phrases can be modified by adverbs. 00:33:23.895 --> 00:33:25.770 align:middle line:84% So you can say "She put them under the tree." 00:33:25.770 --> 00:33:28.080 align:middle line:84% You can also say "She put them right under the tree," 00:33:28.080 --> 00:33:31.590 align:middle line:84% "She put them directly under the tree," where those adverbs are 00:33:31.590 --> 00:33:35.340 align:middle line:84% modifying the prepositional phrase "under the tree," 00:33:35.340 --> 00:33:37.170 align:middle line:84% saying something about the spatial relation 00:33:37.170 --> 00:33:38.590 align:middle line:84% between those things and the tree. 00:33:38.590 --> 00:33:39.090 align:middle line:90% Yeah? 00:33:39.090 --> 00:33:42.720 align:middle line:90% 00:33:42.720 --> 00:33:47.380 align:middle line:84% But we're never going to find a verb that selects 00:33:47.380 --> 00:33:48.610 align:middle line:90% for a prepositional phrase. 00:33:48.610 --> 00:33:51.097 align:middle line:90% 00:33:51.097 --> 00:33:52.930 align:middle line:84% So there are verbs like "depend" that select 00:33:52.930 --> 00:33:55.097 align:middle line:84% for a prepositional phrase where the preposition has 00:33:55.097 --> 00:33:57.520 align:middle line:84% to be something, like "depend" needs "on" We're 00:33:57.520 --> 00:33:59.290 align:middle line:84% never going to find a verb that selects 00:33:59.290 --> 00:34:00.700 align:middle line:84% for a prepositional phrase and it 00:34:00.700 --> 00:34:03.940 align:middle line:84% needs to be modified by a certain kind of modifier. 00:34:03.940 --> 00:34:05.080 align:middle line:90% We'll never find a-- 00:34:05.080 --> 00:34:06.730 align:middle line:90% I made up a verb, "glorf." 00:34:06.730 --> 00:34:07.270 align:middle line:90% Yeah. 00:34:07.270 --> 00:34:10.000 align:middle line:84% And "glorf" says, "I need a prepositional phrase. 00:34:10.000 --> 00:34:12.010 align:middle line:84% I don't care what the preposition is. 00:34:12.010 --> 00:34:14.679 align:middle line:84% You can say 'She glorfed them right under 00:34:14.679 --> 00:34:16.449 align:middle line:84% the tree,' 'She glorfed them right through 00:34:16.449 --> 00:34:18.699 align:middle line:84% the tree,' 'She glorfed them right beside the tree,' I 00:34:18.699 --> 00:34:19.900 align:middle line:84% don't care what the preposition is, 00:34:19.900 --> 00:34:22.460 align:middle line:84% but we need 'right' at the beginning of the prepositional 00:34:22.460 --> 00:34:22.960 align:middle line:90% phrase." 00:34:22.960 --> 00:34:26.230 align:middle line:84% There are no verbs like that in any language. 00:34:26.230 --> 00:34:28.510 align:middle line:90% Selection doesn't work that way. 00:34:28.510 --> 00:34:29.679 align:middle line:90% How does selection work? 00:34:29.679 --> 00:34:31.719 align:middle line:84% If you're selecting for a prepositional phrase, 00:34:31.719 --> 00:34:33.730 align:middle line:84% you're selecting if you have any restrictions 00:34:33.730 --> 00:34:35.449 align:middle line:84% on the kind of prepositional phrase you want. 00:34:35.449 --> 00:34:37.407 align:middle line:84% They are restrictions of the form "I would like 00:34:37.407 --> 00:34:39.288 align:middle line:90% the preposition to be (this)." 00:34:39.288 --> 00:34:41.080 align:middle line:84% Sometimes they are restrictions of the form 00:34:41.080 --> 00:34:42.538 align:middle line:84% "I would like the preposition to be 00:34:42.538 --> 00:34:43.960 align:middle line:90% (this class of prepositions)." 00:34:43.960 --> 00:34:46.389 align:middle line:84% So I might have a slide about this later too. 00:34:46.389 --> 00:34:49.570 align:middle line:84% "Put" can combine with a bunch of different prepositional 00:34:49.570 --> 00:34:51.010 align:middle line:90% phrases. 00:34:51.010 --> 00:34:55.810 align:middle line:84% The preposition can be "under" or "on," "beside," whatever, 00:34:55.810 --> 00:35:00.430 align:middle line:84% but you can't say things like "She put them during the party" 00:35:00.430 --> 00:35:03.010 align:middle line:84% or "She put them despite the rain." 00:35:03.010 --> 00:35:04.510 align:middle line:84% So there are a bunch of prepositions 00:35:04.510 --> 00:35:06.220 align:middle line:90% that don't go well with "put." 00:35:06.220 --> 00:35:08.470 align:middle line:84% It's basically "put" needs a prepositional phrase that 00:35:08.470 --> 00:35:10.450 align:middle line:90% is a location. 00:35:10.450 --> 00:35:12.910 align:middle line:84% So that's a class of prepositions 00:35:12.910 --> 00:35:17.950 align:middle line:84% but it's not every imaginable preposition. 00:35:17.950 --> 00:35:20.280 align:middle line:84% But there are no verbs-- there's no "glorf," 00:35:20.280 --> 00:35:22.780 align:middle line:84% there are no verbs that say, "I want a prepositional phrase. 00:35:22.780 --> 00:35:24.322 align:middle line:84% I don't care what the preposition is, 00:35:24.322 --> 00:35:27.160 align:middle line:84% but I want it to be modified by 'right,'" or "I want it to be 00:35:27.160 --> 00:35:29.410 align:middle line:84% modified by 'directly,'" or even just "I want it to be 00:35:29.410 --> 00:35:30.320 align:middle line:90% modified." 00:35:30.320 --> 00:35:33.860 align:middle line:90% There aren't verbs like that. 00:35:33.860 --> 00:35:37.820 align:middle line:84% So it looks as though heads, like verbs, 00:35:37.820 --> 00:35:40.490 align:middle line:84% get to select for phrases, and they-- this 00:35:40.490 --> 00:35:45.625 align:middle line:84% is an old observation of Noam Chomsky's in a '65 book of his. 00:35:45.625 --> 00:35:47.750 align:middle line:84% He said, looks like there are all these cases where 00:35:47.750 --> 00:35:51.402 align:middle line:84% a verb is not only selecting for a certain kind of phrase, 00:35:51.402 --> 00:35:53.360 align:middle line:84% but it's selecting for a certain kind of phrase 00:35:53.360 --> 00:35:56.600 align:middle line:84% with a particular head or with the head of a particular class 00:35:56.600 --> 00:35:57.580 align:middle line:90% like locative. 00:35:57.580 --> 00:36:01.400 align:middle line:84% I want a locative preposition like "put." 00:36:01.400 --> 00:36:02.570 align:middle line:90% Yeah? 00:36:02.570 --> 00:36:04.570 align:middle line:90% Sound great? 00:36:04.570 --> 00:36:05.480 align:middle line:90% OK. 00:36:05.480 --> 00:36:05.980 align:middle line:90% Cool. 00:36:05.980 --> 00:36:08.850 align:middle line:90% 00:36:08.850 --> 00:36:14.010 align:middle line:84% So once we believe this, once we recognize that this is true, 00:36:14.010 --> 00:36:15.870 align:middle line:84% that this is how selection works, 00:36:15.870 --> 00:36:18.030 align:middle line:84% that verbs get to select for things 00:36:18.030 --> 00:36:20.427 align:middle line:84% they want to merge with-- not just verbs, we'll see. 00:36:20.427 --> 00:36:22.260 align:middle line:84% All kinds of things get to select for things 00:36:22.260 --> 00:36:25.020 align:middle line:90% that they want to merge with. 00:36:25.020 --> 00:36:30.060 align:middle line:84% And if they are picky about the nature of the thing 00:36:30.060 --> 00:36:32.670 align:middle line:84% that they merge with, what they are picky about 00:36:32.670 --> 00:36:36.490 align:middle line:84% is the head of the thing that they merge with, 00:36:36.490 --> 00:36:39.180 align:middle line:84% the thing that the phrase is named after. 00:36:39.180 --> 00:36:43.290 align:middle line:84% So the preposition, in the case of a prepositional phrase. 00:36:43.290 --> 00:36:46.350 align:middle line:84% So once we recognize that that's a thing of the "depend on" 00:36:46.350 --> 00:36:49.920 align:middle line:84% phenomenon, then we can start using it 00:36:49.920 --> 00:36:52.290 align:middle line:84% as a way of looking around for other heads 00:36:52.290 --> 00:36:55.442 align:middle line:84% when we see that kind of relation of pickiness. 00:36:55.442 --> 00:36:57.150 align:middle line:84% If you have this verb then you had better 00:36:57.150 --> 00:36:59.970 align:middle line:90% have this kind of word here. 00:36:59.970 --> 00:37:02.928 align:middle line:84% Then we get to suspect that there's a selection relation 00:37:02.928 --> 00:37:04.470 align:middle line:84% there because now we know that that's 00:37:04.470 --> 00:37:07.510 align:middle line:90% how selection relations work. 00:37:07.510 --> 00:37:10.590 align:middle line:84% So here's another example of the same kind. 00:37:10.590 --> 00:37:12.570 align:middle line:84% "I think that I have won the lottery." 00:37:12.570 --> 00:37:14.280 align:middle line:90% Fine. 00:37:14.280 --> 00:37:17.310 align:middle line:84% That's actually kind of pleasant to think about. 00:37:17.310 --> 00:37:19.670 align:middle line:84% "I wonder whether I have won the lottery." 00:37:19.670 --> 00:37:21.340 align:middle line:90% Also fine. 00:37:21.340 --> 00:37:24.690 align:middle line:84% But "I think whether I have won the lottery." 00:37:24.690 --> 00:37:25.500 align:middle line:90% No. 00:37:25.500 --> 00:37:27.340 align:middle line:90% I cannot say that. 00:37:27.340 --> 00:37:30.390 align:middle line:84% And also "I wonder that I have won the lottery." 00:37:30.390 --> 00:37:32.577 align:middle line:90% Cannot say in modern English. 00:37:32.577 --> 00:37:34.410 align:middle line:84% There are older versions of English in which 00:37:34.410 --> 00:37:35.702 align:middle line:90% you could say things like this. 00:37:35.702 --> 00:37:39.300 align:middle line:84% It meant "I am surprised that I have won the lottery." 00:37:39.300 --> 00:37:42.450 align:middle line:84% But in modern English the first two sentences are OK, 00:37:42.450 --> 00:37:44.550 align:middle line:84% the second two sentences are bad. 00:37:44.550 --> 00:37:45.540 align:middle line:90% Do people agree? 00:37:45.540 --> 00:37:48.120 align:middle line:84% Don't let me run roughshod over your English judgments, 00:37:48.120 --> 00:37:50.200 align:middle line:90% but I think this is true. 00:37:50.200 --> 00:37:51.840 align:middle line:84% So here's another place where there's 00:37:51.840 --> 00:37:54.180 align:middle line:90% this relationship of pickiness. 00:37:54.180 --> 00:37:56.520 align:middle line:84% So we know that if you have a "depends," 00:37:56.520 --> 00:37:58.020 align:middle line:84% you must have a prepositional phrase 00:37:58.020 --> 00:37:59.687 align:middle line:84% and the head of the prepositional phrase 00:37:59.687 --> 00:38:00.840 align:middle line:90% must be "on." 00:38:00.840 --> 00:38:03.330 align:middle line:84% Here we seem to be seeing that with "think" 00:38:03.330 --> 00:38:08.550 align:middle line:84% there can be a clause after it, and then there's a particular 00:38:08.550 --> 00:38:11.340 align:middle line:84% word that starts that clause which must be "that" 00:38:11.340 --> 00:38:14.430 align:middle line:90% and cannot be "whether." 00:38:14.430 --> 00:38:18.150 align:middle line:84% And we're seeing that "wonder" seems 00:38:18.150 --> 00:38:21.220 align:middle line:84% to be selecting for the clause that comes after it, 00:38:21.220 --> 00:38:25.110 align:middle line:84% and it's picky about the properties of the first word 00:38:25.110 --> 00:38:25.840 align:middle line:90% in that clause. 00:38:25.840 --> 00:38:27.673 align:middle line:84% It can be "whether" but it cannot be "that." 00:38:27.673 --> 00:38:29.980 align:middle line:90% 00:38:29.980 --> 00:38:33.360 align:middle line:84% So here we get to make the move that I foreshadowed. 00:38:33.360 --> 00:38:34.300 align:middle line:90% Is that going to go-- 00:38:34.300 --> 00:38:36.390 align:middle line:84% we can say maybe this is like "depend" and "on." 00:38:36.390 --> 00:38:39.030 align:middle line:84% Maybe "think" is selecting for that clause 00:38:39.030 --> 00:38:41.683 align:middle line:84% and maybe that is like "on," so "on" 00:38:41.683 --> 00:38:43.350 align:middle line:84% is the head of the prepositional phrase. 00:38:43.350 --> 00:38:45.405 align:middle line:84% The prepositional phrase is named after "on." 00:38:45.405 --> 00:38:46.800 align:middle line:90% "On" is a preposition. 00:38:46.800 --> 00:38:49.050 align:middle line:84% Similarly, we're going to say this clause has, 00:38:49.050 --> 00:38:51.450 align:middle line:84% as its head, words like "that" and "whether." 00:38:51.450 --> 00:38:55.220 align:middle line:90% 00:38:55.220 --> 00:38:58.400 align:middle line:84% We have a name for words like "that" and "whether." 00:38:58.400 --> 00:39:03.908 align:middle line:84% We call them "complementizers," which-- 00:39:03.908 --> 00:39:06.200 align:middle line:84% I feel as though I've done a lot of apologizing already 00:39:06.200 --> 00:39:08.878 align:middle line:84% for terminology in this class so maybe I'll just stop. 00:39:08.878 --> 00:39:09.920 align:middle line:90% That's what we call them. 00:39:09.920 --> 00:39:11.420 align:middle line:90% We call them complementizers. 00:39:11.420 --> 00:39:14.100 align:middle line:90% Deal. 00:39:14.100 --> 00:39:17.030 align:middle line:84% You may have heard them called other things 00:39:17.030 --> 00:39:18.350 align:middle line:90% in English classes. 00:39:18.350 --> 00:39:21.710 align:middle line:84% "Subordinating conjunctions," they're sometimes called. 00:39:21.710 --> 00:39:23.240 align:middle line:84% Try not to get too hung up on that. 00:39:23.240 --> 00:39:26.570 align:middle line:84% In this class they are called complementizers. 00:39:26.570 --> 00:39:31.610 align:middle line:84% And they have the handy abbreviation C. 00:39:31.610 --> 00:39:39.560 align:middle line:84% So what we'll say is that a verb like "think" selects for CP. 00:39:39.560 --> 00:39:43.400 align:middle line:84% So the name we're going to give to this clause that 00:39:43.400 --> 00:39:47.520 align:middle line:84% comes after "think," "that I have won the lottery," 00:39:47.520 --> 00:39:48.870 align:middle line:90% we're going to call that a CP. 00:39:48.870 --> 00:39:49.370 align:middle line:90% Why? 00:39:49.370 --> 00:39:51.992 align:middle line:84% Well, because we've seen that this relationship of pickiness 00:39:51.992 --> 00:39:54.200 align:middle line:84% holds between "think" and "that" and between "wonder" 00:39:54.200 --> 00:39:57.200 align:middle line:84% and "whether," and that entitles us to believe 00:39:57.200 --> 00:40:00.587 align:middle line:84% "that" and "whether" are the heads of the phrases 00:40:00.587 --> 00:40:02.420 align:middle line:84% that "think" and "wonder" are selecting for. 00:40:02.420 --> 00:40:04.520 align:middle line:84% They're like the "on" in "depend on." 00:40:04.520 --> 00:40:05.270 align:middle line:90% Yeah. 00:40:05.270 --> 00:40:08.160 align:middle line:84% So that's the reason to draw the trees this way. 00:40:08.160 --> 00:40:12.678 align:middle line:84% So we'll say, when you have a clause like "that I have won 00:40:12.678 --> 00:40:14.720 align:middle line:84% the lottery" or "whether I have won the lottery," 00:40:14.720 --> 00:40:16.130 align:middle line:90% that thing is a CP. 00:40:16.130 --> 00:40:18.320 align:middle line:84% "That" or "whether" is its head, and those things 00:40:18.320 --> 00:40:20.540 align:middle line:84% are being selected by verbs like "think" or "wonder." 00:40:20.540 --> 00:40:21.040 align:middle line:90% Yeah. 00:40:21.040 --> 00:40:23.750 align:middle line:90% 00:40:23.750 --> 00:40:25.610 align:middle line:90% OK? 00:40:25.610 --> 00:40:26.210 align:middle line:90% All right. 00:40:26.210 --> 00:40:29.620 align:middle line:90% Those are all complementizers. 00:40:29.620 --> 00:40:30.120 align:middle line:90% OK. 00:40:30.120 --> 00:40:31.245 align:middle line:90% So are we done with syntax? 00:40:31.245 --> 00:40:35.790 align:middle line:90% 00:40:35.790 --> 00:40:37.660 align:middle line:84% We now have this belief in selection 00:40:37.660 --> 00:40:42.870 align:middle line:84% which is allowing us to explain why certain pairs of things 00:40:42.870 --> 00:40:44.680 align:middle line:84% are merging in the way that they are. 00:40:44.680 --> 00:40:48.690 align:middle line:84% So in this example, "I will tickle the child," when 00:40:48.690 --> 00:40:51.900 align:middle line:84% we ask why did you merge "tickle" 00:40:51.900 --> 00:40:54.360 align:middle line:84% together with the noun phrase "the child," 00:40:54.360 --> 00:40:55.950 align:middle line:90% we now have an answer. 00:40:55.950 --> 00:41:00.180 align:middle line:84% We get to say, yeah, we did that because "tickle" 00:41:00.180 --> 00:41:01.960 align:middle line:90% is a transitive verb. 00:41:01.960 --> 00:41:03.840 align:middle line:84% So if we look up "tickle" in the lexicon 00:41:03.840 --> 00:41:05.520 align:middle line:84% the lexicon will tell us, yeah, this 00:41:05.520 --> 00:41:10.290 align:middle line:84% is a verb that at least has a use in which it 00:41:10.290 --> 00:41:13.240 align:middle line:84% wants to combine with a noun phrase object. 00:41:13.240 --> 00:41:17.040 align:middle line:84% So that's why you merged the verb with the noun 00:41:17.040 --> 00:41:18.930 align:middle line:90% phrase "the child." 00:41:18.930 --> 00:41:20.820 align:middle line:84% And similarly, we might say, yeah, 00:41:20.820 --> 00:41:24.320 align:middle line:84% why did we merge the verb phrase together with the T "will"? 00:41:24.320 --> 00:41:25.980 align:middle line:84% And we can say something similar. 00:41:25.980 --> 00:41:28.590 align:middle line:84% The T "will" is selecting for a verb phrase. 00:41:28.590 --> 00:41:32.753 align:middle line:84% It needs a verb phrase as its sister, and so there. 00:41:32.753 --> 00:41:35.170 align:middle line:84% That's why we're merging these two things with each other. 00:41:35.170 --> 00:41:38.730 align:middle line:84% So we have beginners of explanations or answers 00:41:38.730 --> 00:41:41.355 align:middle line:84% to questions like, why did you merge 00:41:41.355 --> 00:41:42.480 align:middle line:90% those things in that order? 00:41:42.480 --> 00:41:44.760 align:middle line:84% Why did you merge those two things together? 00:41:44.760 --> 00:41:46.265 align:middle line:84% The answer is going to sometimes be, 00:41:46.265 --> 00:41:47.640 align:middle line:84% well, because there's a selection 00:41:47.640 --> 00:41:50.500 align:middle line:90% relation between those things. 00:41:50.500 --> 00:41:52.390 align:middle line:90% OK? 00:41:52.390 --> 00:41:54.990 align:middle line:90% But here's the problem. 00:41:54.990 --> 00:41:58.070 align:middle line:84% You can do anything with a feather. 00:41:58.070 --> 00:42:00.342 align:middle line:84% So I spent some time telling you there 00:42:00.342 --> 00:42:02.300 align:middle line:84% are verbs that need objects and there are verbs 00:42:02.300 --> 00:42:03.590 align:middle line:90% that need to not have objects. 00:42:03.590 --> 00:42:05.420 align:middle line:84% There are transitive verbs, there are intransitive verbs, 00:42:05.420 --> 00:42:07.378 align:middle line:84% and OK, there are verbs that can be either one. 00:42:07.378 --> 00:42:11.270 align:middle line:84% And I tried to move us quickly past that class of verbs, 00:42:11.270 --> 00:42:13.980 align:middle line:90% but they exist. 00:42:13.980 --> 00:42:17.360 align:middle line:84% So yeah, the child, we know why the child is in there 00:42:17.360 --> 00:42:19.790 align:middle line:84% and we think that the relationship between "tickle" 00:42:19.790 --> 00:42:22.970 align:middle line:84% and "the child" is a relation of pickiness. 00:42:22.970 --> 00:42:25.940 align:middle line:84% There are some verbs that can merge together with an object 00:42:25.940 --> 00:42:29.940 align:middle line:84% like that and there are other verbs that can't. 00:42:29.940 --> 00:42:32.550 align:middle line:84% But "with a feather" is not like that. 00:42:32.550 --> 00:42:34.260 align:middle line:84% You can tickle a child with a feather. 00:42:34.260 --> 00:42:37.950 align:middle line:84% You can walk down the street with a feather. 00:42:37.950 --> 00:42:39.172 align:middle line:90% You can stand with a feather. 00:42:39.172 --> 00:42:40.380 align:middle line:90% You can dance with a feather. 00:42:40.380 --> 00:42:42.480 align:middle line:84% You can do anything with a feather. 00:42:42.480 --> 00:42:44.433 align:middle line:84% Prepositional phrase "with a feather"-- 00:42:44.433 --> 00:42:46.350 align:middle line:84% you can write your dissertation with a feather 00:42:46.350 --> 00:42:48.188 align:middle line:90% if you work hard at it. 00:42:48.188 --> 00:42:49.980 align:middle line:84% Prepositional phrase like "with a feather," 00:42:49.980 --> 00:42:52.380 align:middle line:90% there's no pickiness involved. 00:42:52.380 --> 00:42:54.210 align:middle line:84% It can just combine with anything. 00:42:54.210 --> 00:42:57.060 align:middle line:90% 00:42:57.060 --> 00:43:01.580 align:middle line:84% So this does not fall under our generalizations 00:43:01.580 --> 00:43:04.250 align:middle line:90% about how selection works. 00:43:04.250 --> 00:43:06.290 align:middle line:84% We don't seem to have the kind of relation 00:43:06.290 --> 00:43:08.780 align:middle line:84% that we have between the verb and the object. 00:43:08.780 --> 00:43:11.030 align:middle line:84% Yeah, there are verbs that want objects and verbs that 00:43:11.030 --> 00:43:12.405 align:middle line:84% don't want objects, but we're not 00:43:12.405 --> 00:43:15.290 align:middle line:84% going to find a verb that needs "with the feather" 00:43:15.290 --> 00:43:18.290 align:middle line:90% in this sense, using a feather. 00:43:18.290 --> 00:43:19.038 align:middle line:90% Yeah? 00:43:19.038 --> 00:43:20.580 align:middle line:84% AUDIENCE: Could you almost look at it 00:43:20.580 --> 00:43:23.080 align:middle line:84% from a perspective of agency, the fact 00:43:23.080 --> 00:43:26.050 align:middle line:84% that you could say "with a feather" but not "with a dog." 00:43:26.050 --> 00:43:28.530 align:middle line:90% The dog is able to-- 00:43:28.530 --> 00:43:31.040 align:middle line:84% you can't tickle a child with a dog. 00:43:31.040 --> 00:43:32.510 align:middle line:90% NORVIN RICHARDS: Yeah. 00:43:32.510 --> 00:43:35.030 align:middle line:90% Well, it's complicated, right? 00:43:35.030 --> 00:43:37.730 align:middle line:84% You can tickle a child with a dog-- well, let's see. 00:43:37.730 --> 00:43:40.640 align:middle line:84% The fact that you cannot tickle a child with a dog, 00:43:40.640 --> 00:43:44.720 align:middle line:84% if it's a fact, I think is probably a fact about dogs. 00:43:44.720 --> 00:43:45.560 align:middle line:90% Right? 00:43:45.560 --> 00:43:51.080 align:middle line:84% I mean, so we're in "colorless green ideas" territory here. 00:43:51.080 --> 00:43:53.660 align:middle line:84% I was about to do the thing that everybody always does when I 00:43:53.660 --> 00:43:57.180 align:middle line:84% show them colorless green ideas where they say, "Well, 00:43:57.180 --> 00:44:00.050 align:middle line:84% but if 'colorless' and 'green' meant different things then 00:44:00.050 --> 00:44:01.760 align:middle line:90% the sentence would be OK." 00:44:01.760 --> 00:44:04.010 align:middle line:84% And I think tickling a child with a dog, 00:44:04.010 --> 00:44:05.990 align:middle line:84% I find myself saying, well, but what if it's 00:44:05.990 --> 00:44:08.090 align:middle line:90% a particularly hairy dog? 00:44:08.090 --> 00:44:12.407 align:middle line:84% And you grab it and rub it on the child 00:44:12.407 --> 00:44:14.990 align:middle line:84% in just the right way, you could imagine maybe being able to-- 00:44:14.990 --> 00:44:15.920 align:middle line:90% but yeah. 00:44:15.920 --> 00:44:19.020 align:middle line:84% "I will tickle the child with the hammer." 00:44:19.020 --> 00:44:21.650 align:middle line:84% Probably you could never tickle the child with a hammer, 00:44:21.650 --> 00:44:26.480 align:middle line:84% but this is 24.900 and we don't have to care about that. 00:44:26.480 --> 00:44:30.290 align:middle line:84% So some things are better tickling instruments 00:44:30.290 --> 00:44:34.700 align:middle line:84% than others, but thankfully I am not here 00:44:34.700 --> 00:44:37.640 align:middle line:84% to educate you about tickling instruments. 00:44:37.640 --> 00:44:40.520 align:middle line:84% That'll be somebody else's job, if anybody. 00:44:40.520 --> 00:44:42.480 align:middle line:90% Yeah. 00:44:42.480 --> 00:44:43.950 align:middle line:84% Not that I'm promising that there 00:44:43.950 --> 00:44:45.705 align:middle line:84% is an MIT class where you can learn 00:44:45.705 --> 00:44:46.830 align:middle line:90% about tickling instruments. 00:44:46.830 --> 00:44:47.910 align:middle line:84% I mean, I suppose there could be, 00:44:47.910 --> 00:44:49.290 align:middle line:84% but it's definitely not this one. 00:44:49.290 --> 00:44:50.190 align:middle line:90% Yeah. 00:44:50.190 --> 00:44:51.370 align:middle line:90% Yeah? 00:44:51.370 --> 00:44:51.870 align:middle line:90% Good. 00:44:51.870 --> 00:44:53.280 align:middle line:90% All right. 00:44:53.280 --> 00:44:55.440 align:middle line:90% OK. 00:44:55.440 --> 00:44:59.850 align:middle line:84% So "with the feather" is another kind of thing. 00:44:59.850 --> 00:45:00.350 align:middle line:90% Yeah. 00:45:00.350 --> 00:45:01.892 align:middle line:84% This is just saying what I just said. 00:45:01.892 --> 00:45:04.550 align:middle line:84% Doesn't seem right to say that "child" or "tickle" selects 00:45:04.550 --> 00:45:08.660 align:middle line:84% "with a feather" because you can do anything with a feather 00:45:08.660 --> 00:45:10.290 align:middle line:90% and, uh-- 00:45:10.290 --> 00:45:10.790 align:middle line:90% there. 00:45:10.790 --> 00:45:11.090 align:middle line:90% Yeah. 00:45:11.090 --> 00:45:13.007 align:middle line:84% So "I will tickle the child with the feather." 00:45:13.007 --> 00:45:15.123 align:middle line:84% "I will devour the child with the feather." 00:45:15.123 --> 00:45:16.790 align:middle line:84% "I will write a novel with the feather." 00:45:16.790 --> 00:45:18.165 align:middle line:84% "I will thrive with the feather." 00:45:18.165 --> 00:45:20.240 align:middle line:84% You can do anything with the feather. 00:45:20.240 --> 00:45:22.890 align:middle line:84% Some of these things are more or less plausible than others, 00:45:22.890 --> 00:45:24.860 align:middle line:90% but they're all grammatical. 00:45:24.860 --> 00:45:25.463 align:middle line:90% Yeah? 00:45:25.463 --> 00:45:26.380 align:middle line:90% Does that sound right? 00:45:26.380 --> 00:45:28.845 align:middle line:90% 00:45:28.845 --> 00:45:30.220 align:middle line:84% I don't know why I used "devour." 00:45:30.220 --> 00:45:35.560 align:middle line:90% 00:45:35.560 --> 00:45:38.720 align:middle line:84% So we said there are verbs like "tickle" and "devour" and 00:45:38.720 --> 00:45:43.760 align:middle line:84% "write" that need to have objects, 00:45:43.760 --> 00:45:45.090 align:middle line:90% or at least can have objects. 00:45:45.090 --> 00:45:46.970 align:middle line:84% And there are other verbs like "thrive" that pickily 00:45:46.970 --> 00:45:48.530 align:middle line:84% can't have objects. so not every verb 00:45:48.530 --> 00:45:49.640 align:middle line:90% can be followed by the child. 00:45:49.640 --> 00:45:50.682 align:middle line:90% I can "tickle the child." 00:45:50.682 --> 00:45:52.040 align:middle line:90% I can "devour the child." 00:45:52.040 --> 00:45:55.280 align:middle line:90% I can "write the child." 00:45:55.280 --> 00:45:59.260 align:middle line:84% But I cannot "thrive the child" no matter how hard I try. 00:45:59.260 --> 00:45:59.875 align:middle line:90% Yeah? 00:45:59.875 --> 00:46:01.250 align:middle line:84% AUDIENCE: In the previous slide-- 00:46:01.250 --> 00:46:01.640 align:middle line:90% NORVIN RICHARDS: Yeah? 00:46:01.640 --> 00:46:03.765 align:middle line:84% AUDIENCE: The fact that with something a little bit 00:46:03.765 --> 00:46:04.880 align:middle line:90% better [INAUDIBLE] verb. 00:46:04.880 --> 00:46:07.172 align:middle line:84% How does that relate to what you said earlier about how 00:46:07.172 --> 00:46:09.140 align:middle line:90% verbs select the preposition? 00:46:09.140 --> 00:46:10.550 align:middle line:90% NORVIN RICHARDS: So yeah, good. 00:46:10.550 --> 00:46:11.570 align:middle line:90% Nice point 00:46:11.570 --> 00:46:13.263 align:middle line:84% AUDIENCE: [INAUDIBLE] doesn't work. 00:46:13.263 --> 00:46:14.180 align:middle line:90% NORVIN RICHARDS: Yeah. 00:46:14.180 --> 00:46:14.888 align:middle line:90% Yeah, Yeah, yeah. 00:46:14.888 --> 00:46:16.050 align:middle line:90% That's a good point. 00:46:16.050 --> 00:46:18.020 align:middle line:84% So maybe this is a way to say it. 00:46:18.020 --> 00:46:23.120 align:middle line:84% If you have a verb and you give that verb everything it needs-- 00:46:23.120 --> 00:46:25.760 align:middle line:84% everything it selects, you can always then 00:46:25.760 --> 00:46:27.710 align:middle line:90% add "with a feather." 00:46:27.710 --> 00:46:32.930 align:middle line:84% So "I will depend on the child with the feather" is fine. 00:46:32.930 --> 00:46:35.300 align:middle line:84% You gave "depend" its prepositional phrase 00:46:35.300 --> 00:46:38.743 align:middle line:84% that it wanted, "on the child," and then 00:46:38.743 --> 00:46:40.160 align:middle line:84% you get to add "with the feather." 00:46:40.160 --> 00:46:41.810 align:middle line:90% "With the feather" is desert. 00:46:41.810 --> 00:46:43.280 align:middle line:84% You never need "with the feather," 00:46:43.280 --> 00:46:45.200 align:middle line:90% but you can always have it. 00:46:45.200 --> 00:46:45.860 align:middle line:90% That's right. 00:46:45.860 --> 00:46:47.580 align:middle line:84% And then yes, you're absolutely right. 00:46:47.580 --> 00:46:49.730 align:middle line:84% You can't say "I will depend with the feather" 00:46:49.730 --> 00:46:51.865 align:middle line:84% because that's a case where "depend" needs 00:46:51.865 --> 00:46:53.990 align:middle line:84% a prepositional phrase and it needs the preposition 00:46:53.990 --> 00:46:54.680 align:middle line:90% to be "on." 00:46:54.680 --> 00:46:56.670 align:middle line:84% But if you give "depend" what it needs, 00:46:56.670 --> 00:46:59.060 align:middle line:84% then you may always add "with the feather." 00:46:59.060 --> 00:47:00.990 align:middle line:90% Yeah. 00:47:00.990 --> 00:47:03.800 align:middle line:90% It's an optional extra, always. 00:47:03.800 --> 00:47:05.030 align:middle line:90% Yeah. 00:47:05.030 --> 00:47:07.910 align:middle line:90% Always, modulo-- "modulo"! 00:47:07.910 --> 00:47:10.010 align:middle line:84% That's a word I learned in grad school. 00:47:10.010 --> 00:47:14.900 align:middle line:84% It means "Always setting aside issues 00:47:14.900 --> 00:47:17.630 align:middle line:84% like, wait, what do you mean you will devour 00:47:17.630 --> 00:47:18.830 align:middle line:90% the child with the feather? 00:47:18.830 --> 00:47:19.910 align:middle line:90% Like, how are you-- what? 00:47:19.910 --> 00:47:20.880 align:middle line:90% How are you going to do that?" 00:47:20.880 --> 00:47:23.213 align:middle line:84% So some of these things it's not so clear what exactly-- 00:47:23.213 --> 00:47:24.770 align:middle line:84% or "I will thrive with the feather." 00:47:24.770 --> 00:47:25.895 align:middle line:90% Maybe that means something. 00:47:25.895 --> 00:47:27.530 align:middle line:84% I will live a long and happy life 00:47:27.530 --> 00:47:29.450 align:middle line:90% as long as I have the feather. 00:47:29.450 --> 00:47:34.700 align:middle line:84% So we're back to "colorless green ideas" territory, 00:47:34.700 --> 00:47:35.450 align:middle line:90% more or less. 00:47:35.450 --> 00:47:38.030 align:middle line:84% We're not in the business of trying to figure out which 00:47:38.030 --> 00:47:39.350 align:middle line:90% of these sentences make sense. 00:47:39.350 --> 00:47:41.225 align:middle line:84% We're in the business of trying to figure out 00:47:41.225 --> 00:47:43.820 align:middle line:84% which of them are grammatical, and they're all grammatical. 00:47:43.820 --> 00:47:45.050 align:middle line:90% That's all I'm asserting. 00:47:45.050 --> 00:47:46.100 align:middle line:90% Yeah? 00:47:46.100 --> 00:47:50.120 align:middle line:84% AUDIENCE: [INAUDIBLE] "...with this feather." 00:47:50.120 --> 00:47:52.545 align:middle line:84% Does "with this feather" modify "the child" or "tickle"? 00:47:52.545 --> 00:47:53.420 align:middle line:90% NORVIN RICHARDS: Yes. 00:47:53.420 --> 00:47:54.588 align:middle line:90% To which the answer is yes. 00:47:54.588 --> 00:47:55.130 align:middle line:90% You're right. 00:47:55.130 --> 00:47:56.637 align:middle line:90% So it's ambiguous. 00:47:56.637 --> 00:47:58.220 align:middle line:84% It can have either of those structures 00:47:58.220 --> 00:47:59.420 align:middle line:90% and either of those meanings. 00:47:59.420 --> 00:47:59.920 align:middle line:90% Yeah. 00:47:59.920 --> 00:48:01.140 align:middle line:90% So good point. 00:48:01.140 --> 00:48:04.275 align:middle line:84% So "I will tickle the child with the feather" could mean-- 00:48:04.275 --> 00:48:05.900 align:middle line:84% I think we talked about this in class-- 00:48:05.900 --> 00:48:07.692 align:middle line:84% I will use the feather to tickle the child, 00:48:07.692 --> 00:48:11.347 align:middle line:84% or I will tickle, possibly with my fingers, the child who 00:48:11.347 --> 00:48:11.930 align:middle line:90% has a feather. 00:48:11.930 --> 00:48:14.180 align:middle line:90% That's absolutely right. 00:48:14.180 --> 00:48:16.130 align:middle line:90% But "with the feather" doesn't-- 00:48:16.130 --> 00:48:19.400 align:middle line:84% so "with the feather," yeah, it's even more adaptable 00:48:19.400 --> 00:48:20.480 align:middle line:90% than I made it sound. 00:48:20.480 --> 00:48:23.900 align:middle line:84% It can combine with noun phrases and with verb phrases, 00:48:23.900 --> 00:48:27.860 align:middle line:84% and it can always combine with noun phrases. 00:48:27.860 --> 00:48:29.252 align:middle line:84% Concentrate on that reading where 00:48:29.252 --> 00:48:30.710 align:middle line:84% it's the child who has the feather, 00:48:30.710 --> 00:48:32.300 align:middle line:84% "I can tickle a child with a feather." 00:48:32.300 --> 00:48:33.853 align:middle line:84% I can tickle an adult with a feather, 00:48:33.853 --> 00:48:35.270 align:middle line:84% I can tickle a dog with a feather, 00:48:35.270 --> 00:48:38.015 align:middle line:84% I can tickle an orange with a feather-- 00:48:38.015 --> 00:48:41.270 align:middle line:84% would be hard, but it's grammatical. 00:48:41.270 --> 00:48:42.170 align:middle line:90% Yeah. 00:48:42.170 --> 00:48:42.950 align:middle line:90% Yeah. 00:48:42.950 --> 00:48:43.490 align:middle line:90% Yeah? 00:48:43.490 --> 00:48:45.907 align:middle line:84% So "with a feather"-- anything can have a feather, and you 00:48:45.907 --> 00:48:47.940 align:middle line:84% can always use a feather to do anything. 00:48:47.940 --> 00:48:50.260 align:middle line:90% That's the moral of all this. 00:48:50.260 --> 00:48:50.760 align:middle line:90% OK? 00:48:50.760 --> 00:48:53.678 align:middle line:90% 00:48:53.678 --> 00:48:54.470 align:middle line:90% People are leaving. 00:48:54.470 --> 00:48:57.236 align:middle line:84% Was it all the talk about devouring children? 00:48:57.236 --> 00:48:58.730 align:middle line:90% AUDIENCE: People are leaving. 00:48:58.730 --> 00:49:00.115 align:middle line:90% NORVIN RICHARDS: Sorry? 00:49:00.115 --> 00:49:01.490 align:middle line:84% AUDIENCE: I was trying to come up 00:49:01.490 --> 00:49:03.748 align:middle line:84% with something that wouldn't depend on this feather. 00:49:03.748 --> 00:49:05.290 align:middle line:84% NORVIN RICHARDS: "With this feather"? 00:49:05.290 --> 00:49:06.190 align:middle line:90% No. 00:49:06.190 --> 00:49:07.300 align:middle line:90% So. 00:49:07.300 --> 00:49:10.810 align:middle line:90% "Who are you with this feather?" 00:49:10.810 --> 00:49:13.220 align:middle line:84% "I am far, far more impressive with this feather." 00:49:13.220 --> 00:49:13.860 align:middle line:90% Yeah. 00:49:13.860 --> 00:49:14.485 align:middle line:90% Yeah. 00:49:14.485 --> 00:49:15.100 align:middle line:90% Yeah? 00:49:15.100 --> 00:49:15.430 align:middle line:90% OK. 00:49:15.430 --> 00:49:15.930 align:middle line:90% All right. 00:49:15.930 --> 00:49:17.800 align:middle line:84% So you can do anything with the feather. 00:49:17.800 --> 00:49:20.560 align:middle line:84% And conversely, so "with a feather" 00:49:20.560 --> 00:49:22.930 align:middle line:84% can combine with anything-- maybe not conversely. 00:49:22.930 --> 00:49:25.630 align:middle line:84% "With a feather" can combine with anything. 00:49:25.630 --> 00:49:28.570 align:middle line:84% Objects can't combine with just anything. 00:49:28.570 --> 00:49:30.870 align:middle line:84% You have to pick a verb that is transitive 00:49:30.870 --> 00:49:32.120 align:middle line:90% or at least can be transitive. 00:49:32.120 --> 00:49:33.378 align:middle line:90% So you can "tickle a child." 00:49:33.378 --> 00:49:34.420 align:middle line:90% You can "devour a child." 00:49:34.420 --> 00:49:36.640 align:middle line:90% You can "write a child." 00:49:36.640 --> 00:49:38.260 align:middle line:90% You cannot "thrive a child." 00:49:38.260 --> 00:49:40.000 align:middle line:84% So if you pick an intransitive verb then 00:49:40.000 --> 00:49:41.990 align:middle line:84% you can't combine it with a direct object. 00:49:41.990 --> 00:49:44.020 align:middle line:84% So there's a selection relation between the verb 00:49:44.020 --> 00:49:44.680 align:middle line:90% and its object. 00:49:44.680 --> 00:49:46.305 align:middle line:84% That's what we were just talking about. 00:49:46.305 --> 00:49:48.550 align:middle line:84% We have to look up verbs like "tickle" and "devour" 00:49:48.550 --> 00:49:50.260 align:middle line:84% and "write" in the lexicon and find out 00:49:50.260 --> 00:49:52.190 align:middle line:84% whether they can combine with objects or not, 00:49:52.190 --> 00:49:53.710 align:middle line:84% but we'll never have to look anything up 00:49:53.710 --> 00:49:55.090 align:middle line:84% to know whether you can add "with a feather." 00:49:55.090 --> 00:49:57.010 align:middle line:84% You can just always add "with a feather." 00:49:57.010 --> 00:49:59.110 align:middle line:90% Yeah? 00:49:59.110 --> 00:50:01.840 align:middle line:84% So yes, when we're building trees 00:50:01.840 --> 00:50:04.450 align:middle line:84% we're going to have selection relations that sometimes tell 00:50:04.450 --> 00:50:06.790 align:middle line:84% us, yeah, if we have, say, this kind of verb, 00:50:06.790 --> 00:50:09.730 align:middle line:84% we'd better have this kind of phrase combining with it. 00:50:09.730 --> 00:50:11.338 align:middle line:90% We have "devour." 00:50:11.338 --> 00:50:12.880 align:middle line:84% It had better combine with an object. 00:50:12.880 --> 00:50:14.583 align:middle line:90% Something has to get devoured. 00:50:14.583 --> 00:50:16.750 align:middle line:84% But then there are going to be other kinds of things 00:50:16.750 --> 00:50:19.790 align:middle line:84% like "with a feather" that are just always options. 00:50:19.790 --> 00:50:22.210 align:middle line:90% You can always put them in. 00:50:22.210 --> 00:50:24.010 align:middle line:90% Another terminology break. 00:50:24.010 --> 00:50:28.010 align:middle line:84% We call things that are selected arguments. 00:50:28.010 --> 00:50:33.550 align:middle line:84% So "the child" is an argument of the verb "tickle," 00:50:33.550 --> 00:50:36.318 align:middle line:84% in this example, and we call phrases 00:50:36.318 --> 00:50:38.110 align:middle line:84% that don't seem to be selected by anything, 00:50:38.110 --> 00:50:39.790 align:middle line:90% we call those adjuncts. 00:50:39.790 --> 00:50:41.590 align:middle line:90% Just the name for those things. 00:50:41.590 --> 00:50:47.430 align:middle line:90% 00:50:47.430 --> 00:50:49.000 align:middle line:90% Here we have to be careful. 00:50:49.000 --> 00:50:52.260 align:middle line:84% This is a point where people get confused sometimes. 00:50:52.260 --> 00:50:55.790 align:middle line:90% So let me see if I can say this. 00:50:55.790 --> 00:50:59.360 align:middle line:84% Knowing me, I will probably say this several different ways 00:50:59.360 --> 00:51:01.580 align:middle line:84% and we'll just see if I manage to say it 00:51:01.580 --> 00:51:04.310 align:middle line:84% in a way that makes it make sense because people 00:51:04.310 --> 00:51:06.140 align:middle line:90% get confused about this. 00:51:06.140 --> 00:51:06.888 align:middle line:90% Arguments. 00:51:06.888 --> 00:51:08.180 align:middle line:90% Here's a way to think about it. 00:51:08.180 --> 00:51:11.060 align:middle line:84% Arguments, like direct objects, arguments 00:51:11.060 --> 00:51:14.310 align:middle line:84% are picky about which heads they can combine with. 00:51:14.310 --> 00:51:18.860 align:middle line:84% So if you're asking, should I put in a noun phrase 00:51:18.860 --> 00:51:20.500 align:middle line:90% object for this verb? 00:51:20.500 --> 00:51:22.250 align:middle line:84% Well, you've got to know what the verb is, 00:51:22.250 --> 00:51:26.030 align:middle line:84% whether it's a transitive verb or an intransitive verb. 00:51:26.030 --> 00:51:27.320 align:middle line:90% Adjuncts are not. 00:51:27.320 --> 00:51:30.110 align:middle line:84% So adjuncts like "with a feather" 00:51:30.110 --> 00:51:33.590 align:middle line:90% can combine with anything. 00:51:33.590 --> 00:51:34.880 align:middle line:90% There's no selection, really. 00:51:34.880 --> 00:51:36.200 align:middle line:84% You don't have to look anything up 00:51:36.200 --> 00:51:38.325 align:middle line:84% to find out whether you can merge "with a feather." 00:51:38.325 --> 00:51:40.950 align:middle line:90% 00:51:40.950 --> 00:51:45.170 align:middle line:84% The confusion comes in like this. 00:51:45.170 --> 00:51:48.030 align:middle line:84% The problem is there are optional arguments. 00:51:48.030 --> 00:51:49.220 align:middle line:90% So I flagged this earlier. 00:51:49.220 --> 00:51:50.830 align:middle line:84% There are verbs that are transitive 00:51:50.830 --> 00:51:52.580 align:middle line:84% and there are verbs that are intransitive, 00:51:52.580 --> 00:51:54.507 align:middle line:84% and in the earlier slides where I 00:51:54.507 --> 00:51:56.090 align:middle line:84% was talking about transitivity I tried 00:51:56.090 --> 00:51:58.200 align:middle line:84% to concentrate on verbs that were 00:51:58.200 --> 00:52:00.830 align:middle line:84% comfortable in their identity as either transitive verbs 00:52:00.830 --> 00:52:03.380 align:middle line:84% or intransitive verbs, but there are many, many verbs 00:52:03.380 --> 00:52:05.465 align:middle line:90% that can be either one. 00:52:05.465 --> 00:52:08.540 align:middle line:84% So you can "dance" or you can "dance a hornpipe." 00:52:08.540 --> 00:52:11.240 align:middle line:90% 00:52:11.240 --> 00:52:12.020 align:middle line:90% Here's the thing. 00:52:12.020 --> 00:52:15.810 align:middle line:90% 00:52:15.810 --> 00:52:18.830 align:middle line:84% So you can "dance" or you can "dance a hornpipe," but when 00:52:18.830 --> 00:52:20.780 align:middle line:84% you "dance a hornpipe," "a hornpipe" 00:52:20.780 --> 00:52:23.420 align:middle line:84% is an argument of "dance" because you 00:52:23.420 --> 00:52:27.260 align:middle line:84% have to know whether the verb can be a transitive verb 00:52:27.260 --> 00:52:29.930 align:middle line:84% to know whether there can be a noun phrase there. 00:52:29.930 --> 00:52:31.380 align:middle line:90% The confusion comes in. 00:52:31.380 --> 00:52:40.180 align:middle line:84% People look at sentences like "I will devour the child 00:52:40.180 --> 00:52:40.930 align:middle line:90% with the feather." 00:52:40.930 --> 00:52:45.330 align:middle line:90% 00:52:45.330 --> 00:52:46.290 align:middle line:90% Let's make it "tickle." 00:52:46.290 --> 00:52:48.233 align:middle line:90% 00:52:48.233 --> 00:52:50.025 align:middle line:84% "I will tickle the child with the feather." 00:52:50.025 --> 00:52:52.650 align:middle line:90% 00:52:52.650 --> 00:52:54.150 align:middle line:84% Syntacticians are the kind of people 00:52:54.150 --> 00:52:56.190 align:middle line:84% for whom "I will devour the child with a feather" 00:52:56.190 --> 00:52:57.435 align:middle line:84% and "I will tickle the child with the feather" 00:52:57.435 --> 00:52:59.130 align:middle line:90% are basically the same sentence. 00:52:59.130 --> 00:53:03.160 align:middle line:84% We're just not interested in the differences between those. 00:53:03.160 --> 00:53:04.890 align:middle line:90% They're basically the same. 00:53:04.890 --> 00:53:06.498 align:middle line:84% People look at that and they go-- 00:53:06.498 --> 00:53:08.040 align:middle line:84% well, maybe I want it to be "devour." 00:53:08.040 --> 00:53:09.332 align:middle line:90% No, we can stick with "tickle." 00:53:09.332 --> 00:53:13.480 align:middle line:90% 00:53:13.480 --> 00:53:16.800 align:middle line:84% So what we say is this is an argument 00:53:16.800 --> 00:53:21.788 align:middle line:90% and this is an adjunct. 00:53:21.788 --> 00:53:24.330 align:middle line:84% To say that this is an argument is to say that whether it can 00:53:24.330 --> 00:53:26.520 align:middle line:84% be there or not is determined by what verb 00:53:26.520 --> 00:53:28.950 align:middle line:84% you've got, whether the verb is transitive or not. 00:53:28.950 --> 00:53:30.325 align:middle line:84% To say that this is an adjunct is 00:53:30.325 --> 00:53:32.617 align:middle line:84% to say that I don't care what the verb is when I decide 00:53:32.617 --> 00:53:33.960 align:middle line:90% whether to put that in or not. 00:53:33.960 --> 00:53:36.270 align:middle line:84% What people get confused by is this. 00:53:36.270 --> 00:53:39.007 align:middle line:84% Adjuncts are pretty typically optional. 00:53:39.007 --> 00:53:40.590 align:middle line:84% You can tickle a child with a feather. 00:53:40.590 --> 00:53:43.980 align:middle line:84% You don't have to, you can just tickle a child. 00:53:43.980 --> 00:53:47.290 align:middle line:84% There's never a requirement that an adjunct be there. 00:53:47.290 --> 00:53:52.142 align:middle line:84% And so this is one of those tests for whether you're 00:53:52.142 --> 00:53:53.850 align:middle line:84% looking at an argument or an adjunct that 00:53:53.850 --> 00:53:56.110 align:middle line:90% only works in one direction. 00:53:56.110 --> 00:54:01.000 align:middle line:84% If it's an adjunct it had better be optional. 00:54:01.000 --> 00:54:04.360 align:middle line:84% But if it's optional, it could be an adjunct 00:54:04.360 --> 00:54:06.460 align:middle line:90% or it could be a hornpipe. 00:54:06.460 --> 00:54:11.560 align:middle line:84% It could be an object, in these kinds of examples, that's 00:54:11.560 --> 00:54:13.060 align:middle line:84% combining with the kind of verb that 00:54:13.060 --> 00:54:15.990 align:middle line:84% can be either transitive or intransitive. 00:54:15.990 --> 00:54:18.810 align:middle line:84% So if you're looking at a phrase and you're saying to yourself 00:54:18.810 --> 00:54:20.460 align:middle line:84% "This phrase is optional," you're 00:54:20.460 --> 00:54:24.503 align:middle line:84% not then entitled to guess that it's an adjunct, necessarily. 00:54:24.503 --> 00:54:25.920 align:middle line:84% You only get to find out that it's 00:54:25.920 --> 00:54:28.470 align:middle line:84% an adjunct by asking yourself, will it 00:54:28.470 --> 00:54:31.420 align:middle line:84% matter what verb I have, let's say? 00:54:31.420 --> 00:54:33.030 align:middle line:84% Or what noun I have, depending on what 00:54:33.030 --> 00:54:34.880 align:middle line:90% you're combining it with. 00:54:34.880 --> 00:54:36.210 align:middle line:90% OK? 00:54:36.210 --> 00:54:37.410 align:middle line:90% Is that clear? 00:54:37.410 --> 00:54:39.750 align:middle line:84% Have I said it in enough different ways, Faith? 00:54:39.750 --> 00:54:41.250 align:middle line:84% AUDIENCE: Is this at all complicated 00:54:41.250 --> 00:54:46.530 align:middle line:84% if you treat the feather as, like, an animate entity? 00:54:46.530 --> 00:54:49.390 align:middle line:84% So if you were to actually [INAUDIBLE] a person 00:54:49.390 --> 00:54:51.876 align:middle line:84% and you would say, I danced with the feather. 00:54:51.876 --> 00:54:53.662 align:middle line:90% Would it mean you're dancing-- 00:54:53.662 --> 00:54:55.167 align:middle line:90% like, is that [INAUDIBLE]? 00:54:55.167 --> 00:54:56.000 align:middle line:90% NORVIN RICHARDS: Oh. 00:54:56.000 --> 00:54:56.450 align:middle line:90% Oh. 00:54:56.450 --> 00:54:56.950 align:middle line:90% Oh. 00:54:56.950 --> 00:54:57.800 align:middle line:90% Oh. 00:54:57.800 --> 00:54:58.650 align:middle line:90% Oh. 00:54:58.650 --> 00:54:59.150 align:middle line:90% Oh. 00:54:59.150 --> 00:54:59.900 align:middle line:90% Oh, dear. 00:54:59.900 --> 00:55:04.100 align:middle line:90% 00:55:04.100 --> 00:55:08.790 align:middle line:84% Well, let's reluctantly put feathers aside for a second 00:55:08.790 --> 00:55:15.300 align:middle line:84% and talk about something that we know is animate, like children. 00:55:15.300 --> 00:55:17.690 align:middle line:90% So you can dance with the child. 00:55:17.690 --> 00:55:20.840 align:middle line:84% You can play games with the child. 00:55:20.840 --> 00:55:23.630 align:middle line:84% You can eat lunch with the child. 00:55:23.630 --> 00:55:26.900 align:middle line:84% "With the child" is probably still 00:55:26.900 --> 00:55:30.180 align:middle line:84% an adjunct in that kind of reading. 00:55:30.180 --> 00:55:32.160 align:middle line:84% So "with the feather"-- you're raising a good 00:55:32.160 --> 00:55:33.810 align:middle line:90% point that with in English-- 00:55:33.810 --> 00:55:37.170 align:middle line:84% I was warning you about Indo-European prepositions. 00:55:37.170 --> 00:55:40.905 align:middle line:84% "With" can mean two subtly different things, right? 00:55:40.905 --> 00:55:43.530 align:middle line:84% So you're pointing out that if I tickle a child with a feather, 00:55:43.530 --> 00:55:45.960 align:middle line:84% I'm using the feather as an instrument for what 00:55:45.960 --> 00:55:47.610 align:middle line:90% I'm going to do. 00:55:47.610 --> 00:55:49.710 align:middle line:90% If I have lunch with a feather-- 00:55:49.710 --> 00:55:52.380 align:middle line:84% if I have lunch with a child, then that "with" 00:55:52.380 --> 00:55:55.140 align:middle line:84% is introducing somebody who is doing something together 00:55:55.140 --> 00:55:56.850 align:middle line:90% with me, right? 00:55:56.850 --> 00:55:59.190 align:middle line:84% We're both doing the activity in the verb frames. 00:55:59.190 --> 00:56:03.630 align:middle line:84% So those are maybe two different "with"s, but they're both ad-- 00:56:03.630 --> 00:56:07.440 align:middle line:84% those two prepositional phrases, they're probably both adjuncts. 00:56:07.440 --> 00:56:10.330 align:middle line:84% And "I will dance with the feather," I guess, 00:56:10.330 --> 00:56:12.280 align:middle line:84% could have either reading in principle. 00:56:12.280 --> 00:56:12.780 align:middle line:90% Yeah. 00:56:12.780 --> 00:56:14.040 align:middle line:90% Yeah. 00:56:14.040 --> 00:56:14.760 align:middle line:90% Yeah. 00:56:14.760 --> 00:56:16.250 align:middle line:90% Yeah? 00:56:16.250 --> 00:56:17.240 align:middle line:90% Cool. 00:56:17.240 --> 00:56:20.280 align:middle line:90% Nice example. 00:56:20.280 --> 00:56:22.080 align:middle line:90% OK. 00:56:22.080 --> 00:56:24.720 align:middle line:84% OK, so that's the distinction between arguments and adjuncts. 00:56:24.720 --> 00:56:27.900 align:middle line:90% 00:56:27.900 --> 00:56:29.612 align:middle line:84% Why are we bothering to distinguish 00:56:29.612 --> 00:56:30.570 align:middle line:90% arguments and adjuncts? 00:56:30.570 --> 00:56:33.070 align:middle line:84% Well, there are various places where they behave differently 00:56:33.070 --> 00:56:35.110 align:middle line:90% in interesting ways. 00:56:35.110 --> 00:56:40.260 align:middle line:84% So take a sentence like "I decided on the boat." 00:56:40.260 --> 00:56:42.270 align:middle line:84% That can mean a couple of different things. 00:56:42.270 --> 00:56:44.412 align:middle line:84% Somebody tell me something that it can mean. 00:56:44.412 --> 00:56:46.380 align:middle line:84% AUDIENCE: I made the decision on the boat. 00:56:46.380 --> 00:56:46.800 align:middle line:90% NORVIN RICHARDS: Sorry? 00:56:46.800 --> 00:56:48.300 align:middle line:84% AUDIENCE: I made the decision while I was on. 00:56:48.300 --> 00:56:48.990 align:middle line:84% NORVIN RICHARDS: "I made the decision 00:56:48.990 --> 00:56:49.620 align:middle line:90% while I was on the boat." 00:56:49.620 --> 00:56:50.520 align:middle line:90% That's one thing it can mean. 00:56:50.520 --> 00:56:51.390 align:middle line:90% What's the other thing? 00:56:51.390 --> 00:56:52.482 align:middle line:84% AUDIENCE: "I was a boat dealer and I 00:56:52.482 --> 00:56:54.107 align:middle line:84% was picking between two different boats 00:56:54.107 --> 00:56:55.430 align:middle line:90% and I decided on this one." 00:56:55.430 --> 00:56:55.800 align:middle line:90% NORVIN RICHARDS: Yeah. 00:56:55.800 --> 00:56:56.508 align:middle line:90% Yeah, yeah, yeah. 00:56:56.508 --> 00:56:59.660 align:middle line:84% So I was deciding between a boat and something else. 00:56:59.660 --> 00:57:01.700 align:middle line:90% You know, am I going to buy-- 00:57:01.700 --> 00:57:03.920 align:middle line:84% or maybe I was deciding between two different boats 00:57:03.920 --> 00:57:05.030 align:middle line:90% and I decided on the boat. 00:57:05.030 --> 00:57:07.580 align:middle line:84% I decided on the big black boat instead 00:57:07.580 --> 00:57:08.690 align:middle line:90% of the small white boat. 00:57:08.690 --> 00:57:09.190 align:middle line:90% Yeah. 00:57:09.190 --> 00:57:09.720 align:middle line:90% Whatever. 00:57:09.720 --> 00:57:10.220 align:middle line:90% Yeah. 00:57:10.220 --> 00:57:12.440 align:middle line:84% Or I decided on the boat rather than the motorcycle. 00:57:12.440 --> 00:57:14.720 align:middle line:84% I was deciding what to spend my lottery money on. 00:57:14.720 --> 00:57:15.500 align:middle line:90% Yeah. 00:57:15.500 --> 00:57:16.190 align:middle line:90% Yeah. 00:57:16.190 --> 00:57:17.648 align:middle line:84% It can mean either of those things. 00:57:17.648 --> 00:57:20.270 align:middle line:90% 00:57:20.270 --> 00:57:21.710 align:middle line:84% Maybe the next slide will do this 00:57:21.710 --> 00:57:24.710 align:middle line:84% and I won't have to mess with the blackboard. 00:57:24.710 --> 00:57:25.550 align:middle line:90% Yes. 00:57:25.550 --> 00:57:29.342 align:middle line:84% So those are the two readings, actually, 00:57:29.342 --> 00:57:30.800 align:middle line:84% in the same order that we got them. 00:57:30.800 --> 00:57:32.240 align:middle line:84% Faith's reading, which is the first one, 00:57:32.240 --> 00:57:34.190 align:middle line:84% I made my decision while I was on the boat. 00:57:34.190 --> 00:57:36.810 align:middle line:84% And then Joseph's reading, I chose the boat. 00:57:36.810 --> 00:57:40.000 align:middle line:84% It can mean either of those things. 00:57:40.000 --> 00:57:41.905 align:middle line:84% Is "on the boat" an argument or an adjunct? 00:57:41.905 --> 00:57:44.850 align:middle line:90% 00:57:44.850 --> 00:57:48.653 align:middle line:84% Given what we've said about arguments and adjuncts. 00:57:48.653 --> 00:57:50.820 align:middle line:84% AUDIENCE: It depends on the meaning that you intend. 00:57:50.820 --> 00:57:52.130 align:middle line:90% NORVIN RICHARDS: Yes. 00:57:52.130 --> 00:57:52.660 align:middle line:90% Yes. 00:57:52.660 --> 00:57:54.212 align:middle line:84% The answer to that question is yes. 00:57:54.212 --> 00:57:55.420 align:middle line:90% Is it an argument or adjunct? 00:57:55.420 --> 00:57:55.920 align:middle line:90% Yes. 00:57:55.920 --> 00:57:56.650 align:middle line:90% Yes, it is. 00:57:56.650 --> 00:57:58.510 align:middle line:90% Yeah. 00:57:58.510 --> 00:58:01.750 align:middle line:84% Suppose we take the first reading for "on the boat." 00:58:01.750 --> 00:58:04.240 align:middle line:84% So I made my decision while I was on the boat. 00:58:04.240 --> 00:58:06.220 align:middle line:84% Is that an argument or an adjunct, 00:58:06.220 --> 00:58:08.870 align:middle line:90% that version of on the boat? 00:58:08.870 --> 00:58:09.780 align:middle line:90% AUDIENCE: Adjunct. 00:58:09.780 --> 00:58:11.155 align:middle line:84% NORVIN RICHARDS: It's an adjunct. 00:58:11.155 --> 00:58:13.541 align:middle line:90% You can do anything on a boat. 00:58:13.541 --> 00:58:14.300 align:middle line:90% Sleep on a boat. 00:58:14.300 --> 00:58:16.520 align:middle line:84% Write your dissertation on a boat. 00:58:16.520 --> 00:58:18.500 align:middle line:84% Lots of things you can do on a boat. 00:58:18.500 --> 00:58:21.260 align:middle line:84% On a boat can combine with all kinds of things. 00:58:21.260 --> 00:58:24.470 align:middle line:84% But the second reading where I decided on the boat 00:58:24.470 --> 00:58:28.310 align:middle line:84% ends up meaning "I chose the boat," that's 00:58:28.310 --> 00:58:31.970 align:middle line:84% kind of an idiosyncratic fact about "decide" and "on," right? 00:58:31.970 --> 00:58:35.900 align:middle line:84% That "decide" and "on" can squash together to mean 00:58:35.900 --> 00:58:37.790 align:middle line:90% "choose." 00:58:37.790 --> 00:58:40.040 align:middle line:90% Right? 00:58:40.040 --> 00:58:42.860 align:middle line:84% You can't do this even with other expressions that 00:58:42.860 --> 00:58:44.270 align:middle line:90% more or less mean "decide." 00:58:44.270 --> 00:58:47.000 align:middle line:84% You can't say "I made up my mind on the boat" 00:58:47.000 --> 00:58:50.220 align:middle line:84% and mean "I chose the boat," I don't think. 00:58:50.220 --> 00:58:50.720 align:middle line:90% Right? 00:58:50.720 --> 00:58:53.810 align:middle line:90% 00:58:53.810 --> 00:58:55.820 align:middle line:84% And you certainly can't combine "on the boat" 00:58:55.820 --> 00:58:58.550 align:middle line:84% with any random verb and expect to get a new meaning. 00:58:58.550 --> 00:59:01.890 align:middle line:90% That's not the way life works. 00:59:01.890 --> 00:59:04.940 align:middle line:84% So there's an adjunct meaning of "on the boat" 00:59:04.940 --> 00:59:08.210 align:middle line:84% and there's an argument meaning of "on the boat." 00:59:08.210 --> 00:59:09.080 align:middle line:90% OK. 00:59:09.080 --> 00:59:09.650 align:middle line:90% Cool. 00:59:09.650 --> 00:59:12.140 align:middle line:84% Now, yeah, is it an argument or adjunct? 00:59:12.140 --> 00:59:12.920 align:middle line:90% Yes. 00:59:12.920 --> 00:59:15.170 align:middle line:84% So the first one is an adjunct, the second one is an argument. 00:59:15.170 --> 00:59:15.670 align:middle line:90% Good. 00:59:15.670 --> 00:59:18.070 align:middle line:90% 00:59:18.070 --> 00:59:20.110 align:middle line:84% Consider a sentence like, "I decided on the boat 00:59:20.110 --> 00:59:21.730 align:middle line:90% on the plane." 00:59:21.730 --> 00:59:23.830 align:middle line:84% Please, while you're considering the sentence, 00:59:23.830 --> 00:59:26.560 align:middle line:84% do not think about cases where boats are on planes or planes 00:59:26.560 --> 00:59:27.400 align:middle line:90% are on boats. 00:59:27.400 --> 00:59:29.110 align:middle line:84% There are planes that are also boats-- 00:59:29.110 --> 00:59:32.770 align:middle line:84% I've been teaching this class to MIT undergrads for a while. 00:59:32.770 --> 00:59:34.960 align:middle line:84% I know the kinds of moves that you 00:59:34.960 --> 00:59:37.150 align:middle line:90% are thinking about even now. 00:59:37.150 --> 00:59:40.510 align:middle line:90% So just consider cases where-- 00:59:40.510 --> 00:59:43.480 align:middle line:84% so the imaginable things that it could mean given those 00:59:43.480 --> 00:59:45.970 align:middle line:90% two readings for "on the boat." 00:59:45.970 --> 00:59:47.020 align:middle line:90% I think there are two. 00:59:47.020 --> 00:59:52.930 align:middle line:84% It could mean, "While I was on the plane I chose the boat," 00:59:52.930 --> 00:59:57.800 align:middle line:84% or it could mean "While I was on the boat I chose the plane." 00:59:57.800 --> 00:59:58.640 align:middle line:90% Yeah? 00:59:58.640 --> 01:00:02.230 align:middle line:84% In principle it ought to be able to mean either of those things. 01:00:02.230 --> 01:00:05.835 align:middle line:84% Can it mean both of those things? 01:00:05.835 --> 01:00:11.280 align:middle line:84% AUDIENCE: [INAUDIBLE],, but I just-- 01:00:11.280 --> 01:00:14.080 align:middle line:84% I decided on the boat on the plane. 01:00:14.080 --> 01:00:18.133 align:middle line:84% Like, deciding on the plane on the boat sounds much weirder. 01:00:18.133 --> 01:00:19.050 align:middle line:90% NORVIN RICHARDS: Yeah. 01:00:19.050 --> 01:00:19.680 align:middle line:90% Here, I'll tell you what. 01:00:19.680 --> 01:00:21.222 align:middle line:84% Let me write these readings down just 01:00:21.222 --> 01:00:24.918 align:middle line:84% so we can point at them as we're talking because otherwise I 01:00:24.918 --> 01:00:26.460 align:middle line:84% can tell we're going to get confused. 01:00:26.460 --> 01:00:29.230 align:middle line:84% I'm already confused and we've only just started. 01:00:29.230 --> 01:00:39.984 align:middle line:84% So one reading is "I chose the boat while I was on the plane." 01:00:39.984 --> 01:00:51.940 align:middle line:84% The other is "I chose the plane while I was on the boat." 01:00:51.940 --> 01:00:57.780 align:middle line:90% 01:00:57.780 --> 01:00:59.460 align:middle line:90% Is that legible at all? 01:00:59.460 --> 01:01:00.740 align:middle line:90% Can you read that, any of you? 01:01:00.740 --> 01:01:01.240 align:middle line:90% Yes? 01:01:01.240 --> 01:01:01.610 align:middle line:90% Sort of? 01:01:01.610 --> 01:01:02.110 align:middle line:90% Kind of? 01:01:02.110 --> 01:01:03.360 align:middle line:90% OK, good. 01:01:03.360 --> 01:01:04.800 align:middle line:90% So those are the two readings. 01:01:04.800 --> 01:01:06.638 align:middle line:84% Call them reading one and reading two, 01:01:06.638 --> 01:01:08.430 align:middle line:84% or we could call them the "choose the boat" 01:01:08.430 --> 01:01:10.222 align:middle line:84% reading and the "choose the plane" reading. 01:01:10.222 --> 01:01:12.555 align:middle line:84% And now we're trying to figure out which of those things 01:01:12.555 --> 01:01:13.140 align:middle line:90% can this mean. 01:01:13.140 --> 01:01:13.640 align:middle line:90% Kate? 01:01:13.640 --> 01:01:14.182 align:middle line:90% AUDIENCE: OK. 01:01:14.182 --> 01:01:19.198 align:middle line:84% So instinctively I chose reading one because the boat just-- 01:01:19.198 --> 01:01:19.740 align:middle line:90% I don't know. 01:01:19.740 --> 01:01:21.720 align:middle line:90% That feels like it makes sense. 01:01:21.720 --> 01:01:25.420 align:middle line:84% But then I thought, if you work hard enough to make sense, 01:01:25.420 --> 01:01:28.143 align:middle line:84% especially if you were to replace 01:01:28.143 --> 01:01:30.060 align:middle line:84% the second prepositional phrase with something 01:01:30.060 --> 01:01:32.550 align:middle line:90% more simple, like "at lunch"-- 01:01:32.550 --> 01:01:33.540 align:middle line:90% NORVIN RICHARDS: Oh. 01:01:33.540 --> 01:01:36.880 align:middle line:84% AUDIENCE: In [? formal, ?] or yesterday, I guess, 01:01:36.880 --> 01:01:37.560 align:middle line:90% which isn't-- 01:01:37.560 --> 01:01:38.102 align:middle line:90% I don't know. 01:01:38.102 --> 01:01:39.016 align:middle line:90% [INAUDIBLE] 01:01:39.016 --> 01:01:42.980 align:middle line:84% But "I decided on the boat at lunch" makes sense. 01:01:42.980 --> 01:01:46.065 align:middle line:84% You still have essentially the same content. 01:01:46.065 --> 01:01:46.940 align:middle line:90% NORVIN RICHARDS: Yes. 01:01:46.940 --> 01:01:48.320 align:middle line:90% Yes. 01:01:48.320 --> 01:01:51.530 align:middle line:90% So that's a nice point. 01:01:51.530 --> 01:01:55.880 align:middle line:84% Let me react to it by attempting to squash it. 01:01:55.880 --> 01:01:59.900 align:middle line:84% So don't change either of the prepositional phrases 01:01:59.900 --> 01:02:00.878 align:middle line:90% into anything else. 01:02:00.878 --> 01:02:02.420 align:middle line:84% But we are going to want to come back 01:02:02.420 --> 01:02:05.670 align:middle line:84% to that because that's a nice point that you're making. 01:02:05.670 --> 01:02:12.590 align:middle line:84% I think if I say "I decided on the boat at lunch," 01:02:12.590 --> 01:02:14.930 align:middle line:84% that we're back to the original ambiguity. 01:02:14.930 --> 01:02:16.100 align:middle line:90% Is that your feeling? 01:02:16.100 --> 01:02:18.560 align:middle line:84% That that can mean either I chose the boat at lunch 01:02:18.560 --> 01:02:24.380 align:middle line:84% or I decided while I was on the boat at lunch. 01:02:24.380 --> 01:02:25.910 align:middle line:90% Is that true, do you think? 01:02:25.910 --> 01:02:26.993 align:middle line:90% AUDIENCE: I just got lost. 01:02:26.993 --> 01:02:27.993 align:middle line:90% NORVIN RICHARDS: Me too. 01:02:27.993 --> 01:02:28.520 align:middle line:90% Me too. 01:02:28.520 --> 01:02:30.680 align:middle line:84% So yeah, stick to these prepositions 01:02:30.680 --> 01:02:32.000 align:middle line:90% or these prepositions. 01:02:32.000 --> 01:02:33.708 align:middle line:84% That sentence, we're trying to figure out 01:02:33.708 --> 01:02:34.970 align:middle line:90% what that sentence can mean. 01:02:34.970 --> 01:02:35.970 align:middle line:90% What can that sentence mean? 01:02:35.970 --> 01:02:36.470 align:middle line:90% Yeah? 01:02:36.470 --> 01:02:38.057 align:middle line:84% AUDIENCE: I think the first one would 01:02:38.057 --> 01:02:41.950 align:middle line:84% be more usual in the sense that if you tried to-- 01:02:41.950 --> 01:02:44.592 align:middle line:84% I mean, I know you told us not to replace anything, but-- 01:02:44.592 --> 01:02:45.925 align:middle line:90% NORVIN RICHARDS: It's all right. 01:02:45.925 --> 01:02:49.180 align:middle line:84% AUDIENCE: If you do try to replace "on the boat" 01:02:49.180 --> 01:02:53.290 align:middle line:84% with another adjunct, I don't know, like "the building," 01:02:53.290 --> 01:02:55.180 align:middle line:90% it would be a little weird. 01:02:55.180 --> 01:02:57.130 align:middle line:90% It would pass off as something-- 01:02:57.130 --> 01:02:59.470 align:middle line:84% I would pass it off as something that you could say, 01:02:59.470 --> 01:03:02.410 align:middle line:84% but I wouldn't quite say it and it sounds super weird to say. 01:03:02.410 --> 01:03:05.980 align:middle line:84% Like, "I decided in the building on the plane"? 01:03:05.980 --> 01:03:09.590 align:middle line:84% Is this clearly deciding on the plane? 01:03:09.590 --> 01:03:13.030 align:middle line:84% Decide on the plane, but it sounds weird 01:03:13.030 --> 01:03:15.075 align:middle line:84% because "decided on" should be kind of better. 01:03:15.075 --> 01:03:16.450 align:middle line:84% NORVIN RICHARDS: OK, so now we've 01:03:16.450 --> 01:03:18.700 align:middle line:84% had two votes for changing some of the prepositions, 01:03:18.700 --> 01:03:20.780 align:middle line:90% so I'll stop fighting it. 01:03:20.780 --> 01:03:31.510 align:middle line:84% We could also be thinking about, "I decided in the building 01:03:31.510 --> 01:03:32.170 align:middle line:90% on the boat." 01:03:32.170 --> 01:03:38.420 align:middle line:90% 01:03:38.420 --> 01:03:41.150 align:middle line:90% Is this grammatical? 01:03:41.150 --> 01:03:42.566 align:middle line:90% Sort of? 01:03:42.566 --> 01:03:43.490 align:middle line:90% Yeah? 01:03:43.490 --> 01:03:44.270 align:middle line:90% Yeah? 01:03:44.270 --> 01:03:48.597 align:middle line:84% And does this mean I chose the boat 01:03:48.597 --> 01:03:49.805 align:middle line:90% while it was in the building? 01:03:49.805 --> 01:03:53.010 align:middle line:90% 01:03:53.010 --> 01:03:53.760 align:middle line:90% It can. 01:03:53.760 --> 01:03:54.570 align:middle line:90% Yeah. 01:03:54.570 --> 01:03:57.030 align:middle line:84% I guess it's hard for me to be both in a building 01:03:57.030 --> 01:04:00.120 align:middle line:84% and on a boat unless the building is on the boat 01:04:00.120 --> 01:04:01.430 align:middle line:90% or the boat is on the building. 01:04:01.430 --> 01:04:02.490 align:middle line:90% Yeah. 01:04:02.490 --> 01:04:03.510 align:middle line:90% Many, many hands. 01:04:03.510 --> 01:04:04.665 align:middle line:90% Yes? 01:04:04.665 --> 01:04:05.290 align:middle line:90% AUDIENCE: Yeah. 01:04:05.290 --> 01:04:08.410 align:middle line:90% 01:04:08.410 --> 01:04:13.340 align:middle line:84% My thinking was that when the prepositional phrase is 01:04:13.340 --> 01:04:19.010 align:middle line:84% in argument, it's more attached to the verb by association. 01:04:19.010 --> 01:04:25.890 align:middle line:84% So it kind of makes less sense to split in the middle. 01:04:25.890 --> 01:04:30.940 align:middle line:84% So that sentence that you put below there 01:04:30.940 --> 01:04:38.170 align:middle line:84% doesn't quite go as well, and so that's 01:04:38.170 --> 01:04:44.017 align:middle line:84% why choosing the boat on the plane makes more sense. 01:04:44.017 --> 01:04:44.850 align:middle line:90% NORVIN RICHARDS: OK. 01:04:44.850 --> 01:04:45.350 align:middle line:90% Cool. 01:04:45.350 --> 01:04:48.450 align:middle line:84% So actually, what you just said and what Faith just said 01:04:48.450 --> 01:04:50.520 align:middle line:84% and what you just said a second ago, 01:04:50.520 --> 01:04:52.497 align:middle line:90% these dovetail nicely, I think. 01:04:52.497 --> 01:04:54.330 align:middle line:84% Let me now try to summarize something that I 01:04:54.330 --> 01:04:57.240 align:middle line:90% think all three of you said. 01:04:57.240 --> 01:04:59.670 align:middle line:90% Maybe we can say it like this. 01:04:59.670 --> 01:05:01.830 align:middle line:84% When I asked, can this mean "I chose the boat 01:05:01.830 --> 01:05:03.330 align:middle line:90% while it was in the building"? 01:05:03.330 --> 01:05:06.150 align:middle line:90% Everybody was like, nah. 01:05:06.150 --> 01:05:08.970 align:middle line:84% I think it can, but for me I have to pronounce it 01:05:08.970 --> 01:05:10.630 align:middle line:90% in a particular way. 01:05:10.630 --> 01:05:17.430 align:middle line:84% So I can say "I decided, in the building, on the boat." 01:05:17.430 --> 01:05:21.090 align:middle line:84% I think that's the easiest way for me to say it and have 01:05:21.090 --> 01:05:23.280 align:middle line:90% it mean that. 01:05:23.280 --> 01:05:26.010 align:middle line:84% So I did something fancy with my voice 01:05:26.010 --> 01:05:29.250 align:middle line:84% there that involved trying to hide this prepositional phrase. 01:05:29.250 --> 01:05:30.810 align:middle line:90% Yeah. 01:05:30.810 --> 01:05:34.200 align:middle line:84% It's sometimes called a parenthetical 01:05:34.200 --> 01:05:36.780 align:middle line:84% where you put in these things in strange places. 01:05:36.780 --> 01:05:40.170 align:middle line:84% There's lots of interesting work on parentheticals. 01:05:40.170 --> 01:05:42.600 align:middle line:84% So one thing that all of you are teaching us, 01:05:42.600 --> 01:05:44.250 align:middle line:84% maybe, is that it's dangerous for me 01:05:44.250 --> 01:05:46.830 align:middle line:84% to just show you these words on this slide and ask you 01:05:46.830 --> 01:05:47.610 align:middle line:90% what they mean. 01:05:47.610 --> 01:05:50.130 align:middle line:84% What I should really do is pronounce the sentence at you. 01:05:50.130 --> 01:05:53.220 align:middle line:84% So when you're asking yourself how many things can this mean, 01:05:53.220 --> 01:05:57.210 align:middle line:84% don't ask yourself what "I decided, on the boat, 01:05:57.210 --> 01:05:59.280 align:middle line:90% on the plane" means. 01:05:59.280 --> 01:06:02.160 align:middle line:84% Ask yourself, what does "I decided on the boat 01:06:02.160 --> 01:06:05.980 align:middle line:90% on the plane" mean? 01:06:05.980 --> 01:06:07.710 align:middle line:90% Should I do that again? 01:06:07.710 --> 01:06:11.400 align:middle line:84% How many of you think that it can mean number one? 01:06:11.400 --> 01:06:13.920 align:middle line:84% "I decided on the boat on the plane." 01:06:13.920 --> 01:06:16.918 align:middle line:84% How many of you think that it can mean number two? 01:06:16.918 --> 01:06:19.210 align:middle line:84% Is there anybody who thinks that it can mean number two 01:06:19.210 --> 01:06:20.968 align:middle line:90% and cannot mean number one? 01:06:20.968 --> 01:06:23.260 align:middle line:84% Is there anybody who thinks that it can mean number one 01:06:23.260 --> 01:06:25.690 align:middle line:90% and cannot mean number two? 01:06:25.690 --> 01:06:26.560 align:middle line:90% OK. 01:06:26.560 --> 01:06:29.477 align:middle line:84% So through the power of democracy 01:06:29.477 --> 01:06:31.810 align:middle line:84% we've come to the conclusion that it can mean number one 01:06:31.810 --> 01:06:34.810 align:middle line:84% and can't mean number two, except for a couple of you who 01:06:34.810 --> 01:06:36.730 align:middle line:84% feel that it can mean either one. 01:06:36.730 --> 01:06:37.330 align:middle line:90% Yeah? 01:06:37.330 --> 01:06:39.790 align:middle line:84% I think the ones who feel that it can mean either one 01:06:39.790 --> 01:06:43.420 align:middle line:84% are ones who maybe have ways of doing parentheticals that 01:06:43.420 --> 01:06:47.560 align:middle line:84% are less dramatic than I just did, this process of hiding 01:06:47.560 --> 01:06:49.480 align:middle line:90% a prepositional phrase. 01:06:49.480 --> 01:06:52.390 align:middle line:84% Because I agree that you can say "I decided"-- 01:06:52.390 --> 01:06:53.380 align:middle line:90% see if you agree-- 01:06:53.380 --> 01:06:57.250 align:middle line:84% "I decided on the boat on the plane" 01:06:57.250 --> 01:07:00.100 align:middle line:84% can mean "I chose the plane while I was on the boat." 01:07:00.100 --> 01:07:02.770 align:middle line:84% I have to say that in a particular way, 01:07:02.770 --> 01:07:05.888 align:middle line:84% play those games with pitch as I speak. 01:07:05.888 --> 01:07:07.930 align:middle line:84% Lots of interesting questions about what the heck 01:07:07.930 --> 01:07:09.298 align:middle line:90% I am doing when I do that. 01:07:09.298 --> 01:07:10.840 align:middle line:84% Yeah, you had a question a while ago. 01:07:10.840 --> 01:07:11.470 align:middle line:90% Sorry. 01:07:11.470 --> 01:07:11.970 align:middle line:90% Yep, you. 01:07:11.970 --> 01:07:14.053 align:middle line:84% AUDIENCE: What do you think about the sentence "I, 01:07:14.053 --> 01:07:15.700 align:middle line:84% in the building, decided on the boat." 01:07:15.700 --> 01:07:16.720 align:middle line:90% NORVIN RICHARDS: Whoa. 01:07:16.720 --> 01:07:21.000 align:middle line:84% "I, in the building, decided on the boat." 01:07:21.000 --> 01:07:22.665 align:middle line:90% You asked this question before. 01:07:22.665 --> 01:07:25.170 align:middle line:84% You're the guy who likes to modify pronouns 01:07:25.170 --> 01:07:27.300 align:middle line:90% with prepositional phrases. 01:07:27.300 --> 01:07:29.103 align:middle line:90% What's your name? 01:07:29.103 --> 01:07:31.270 align:middle line:84% AUDIENCE: I'm the guy who likes to modify pronouns-- 01:07:31.270 --> 01:07:32.853 align:middle line:84% NORVIN RICHARDS: OK, that's your name. 01:07:32.853 --> 01:07:33.880 align:middle line:90% OK, the guy who likes-- 01:07:33.880 --> 01:07:35.170 align:middle line:90% do you have a nickname? 01:07:35.170 --> 01:07:35.670 align:middle line:90% "Guy." 01:07:35.670 --> 01:07:38.530 align:middle line:90% We'll call you "Guy." 01:07:38.530 --> 01:07:39.030 align:middle line:90% Cool. 01:07:39.030 --> 01:07:44.670 align:middle line:84% So I, in the building, decided on the boat. 01:07:44.670 --> 01:07:46.270 align:middle line:90% Hey, you think I could do that? 01:07:46.270 --> 01:07:46.770 align:middle line:90% Yeah? 01:07:46.770 --> 01:07:50.874 align:middle line:90% 01:07:50.874 --> 01:07:54.380 align:middle line:84% Then we have the problem that if I'm in the building 01:07:54.380 --> 01:07:58.430 align:middle line:84% I can't also be on a boat unless really unlikely things are 01:07:58.430 --> 01:07:59.690 align:middle line:90% happening. 01:07:59.690 --> 01:08:00.290 align:middle line:90% Right? 01:08:00.290 --> 01:08:01.550 align:middle line:84% Unless there's a building on a boat 01:08:01.550 --> 01:08:03.600 align:middle line:84% or a boat on a building or something like that. 01:08:03.600 --> 01:08:06.017 align:middle line:84% And so you're kind of biased in favor of the reading where 01:08:06.017 --> 01:08:07.730 align:middle line:84% I chose the boat because we've also 01:08:07.730 --> 01:08:10.010 align:middle line:84% said that I'm in the building, thanks 01:08:10.010 --> 01:08:14.210 align:middle line:84% to the prepositional phrase that you're named after, 01:08:14.210 --> 01:08:16.689 align:middle line:84% the one that modifies the pronoun. 01:08:16.689 --> 01:08:18.189 align:middle line:90% That's a nice example. 01:08:18.189 --> 01:08:18.689 align:middle line:90% Faith? 01:08:18.689 --> 01:08:20.064 align:middle line:84% AUDIENCE: Don't we have something 01:08:20.064 --> 01:08:23.095 align:middle line:84% similar that happens if you say, "I decided to run on the boat?" 01:08:23.095 --> 01:08:24.880 align:middle line:84% Because either you're running on the boat 01:08:24.880 --> 01:08:28.300 align:middle line:84% or you're deciding to run for president-- 01:08:28.300 --> 01:08:29.193 align:middle line:90% I don't know. 01:08:29.193 --> 01:08:30.700 align:middle line:84% Campaign while you're on the boat. 01:08:30.700 --> 01:08:31.015 align:middle line:90% NORVIN RICHARDS: Yeah. 01:08:31.015 --> 01:08:31.330 align:middle line:90% Yeah. 01:08:31.330 --> 01:08:32.622 align:middle line:90% "I decided to run on the boat." 01:08:32.622 --> 01:08:33.665 align:middle line:90% That's a nice example. 01:08:33.665 --> 01:08:35.290 align:middle line:84% Yeah, so I decided to run for president 01:08:35.290 --> 01:08:37.120 align:middle line:84% or I decided to be in the marathon 01:08:37.120 --> 01:08:40.603 align:middle line:84% while I was on the boat, or I decided, 01:08:40.603 --> 01:08:41.770 align:middle line:90% "Here's a fun thing I'll do. 01:08:41.770 --> 01:08:42.687 align:middle line:90% I'll run on the boat." 01:08:42.687 --> 01:08:44.050 align:middle line:90% Yeah. 01:08:44.050 --> 01:08:44.590 align:middle line:90% Right. 01:08:44.590 --> 01:08:46.833 align:middle line:90% Those are good examples where-- 01:08:46.833 --> 01:08:48.625 align:middle line:84% actually, all of these are examples where-- 01:08:48.625 --> 01:08:51.460 align:middle line:90% 01:08:51.460 --> 01:08:52.910 align:middle line:90% yeah, let me back up. 01:08:52.910 --> 01:08:54.642 align:middle line:84% So "I decided on the boat on the plane." 01:08:54.642 --> 01:08:57.100 align:middle line:84% Through the power of democracy we've come to the conclusion 01:08:57.100 --> 01:08:59.380 align:middle line:90% that that means number one. 01:08:59.380 --> 01:09:02.830 align:middle line:84% What that means is there are two places 01:09:02.830 --> 01:09:05.740 align:middle line:84% for on the boat for prepositional phrases to be. 01:09:05.740 --> 01:09:08.151 align:middle line:84% There's the place where it gets to be an argument 01:09:08.151 --> 01:09:10.359 align:middle line:84% and there's the place where it gets to be an adjunct. 01:09:10.359 --> 01:09:11.515 align:middle line:90% Right? 01:09:11.515 --> 01:09:13.390 align:middle line:84% And the place where it gets to be an argument 01:09:13.390 --> 01:09:14.457 align:middle line:90% is closer to the verb. 01:09:14.457 --> 01:09:15.790 align:middle line:90% Several of you said this, right? 01:09:15.790 --> 01:09:18.207 align:middle line:84% So if it's right after the verb it gets to be an argument, 01:09:18.207 --> 01:09:20.500 align:middle line:84% and if it's not right after the verb then 01:09:20.500 --> 01:09:21.970 align:middle line:90% it needs to be an adjunct. 01:09:21.970 --> 01:09:25.000 align:middle line:84% And in simple sentences, like "I decided on the boat," 01:09:25.000 --> 01:09:26.442 align:middle line:90% you don't know where it is. 01:09:26.442 --> 01:09:28.359 align:middle line:84% So you don't know whether it's in the argument 01:09:28.359 --> 01:09:31.060 align:middle line:84% place or the adjunct place, and so the example is ambiguous. 01:09:31.060 --> 01:09:33.430 align:middle line:84% It's kind of like the ambiguity of "I will tickle 01:09:33.430 --> 01:09:35.582 align:middle line:84% the child with the feather," where you don't know 01:09:35.582 --> 01:09:37.540 align:middle line:84% whether "with the feather" modifies "the child" 01:09:37.540 --> 01:09:40.210 align:middle line:84% or the verb phrase "tickled the child." 01:09:40.210 --> 01:09:43.300 align:middle line:84% Or, and now this is why finally I'm 01:09:43.300 --> 01:09:45.700 align:middle line:84% becoming relevant to what you just said, Faith, 01:09:45.700 --> 01:09:47.859 align:middle line:84% that's another example where "on the boat," 01:09:47.859 --> 01:09:49.210 align:middle line:90% it's ambiguous where you put. 01:09:49.210 --> 01:09:51.460 align:middle line:84% It could be a couple different places in the structure 01:09:51.460 --> 01:09:54.100 align:middle line:84% and it could modify different things. 01:09:54.100 --> 01:09:54.640 align:middle line:90% Yeah. 01:09:54.640 --> 01:09:56.650 align:middle line:84% So we're getting these ambiguities. 01:09:56.650 --> 01:09:58.600 align:middle line:84% And the fact that these ambiguities disappear 01:09:58.600 --> 01:10:00.580 align:middle line:84% if you pile these prepositional phrases up 01:10:00.580 --> 01:10:01.870 align:middle line:90% disappear for most of us. 01:10:01.870 --> 01:10:05.627 align:middle line:84% Some of you are more creative with prepositional phrases. 01:10:05.627 --> 01:10:07.960 align:middle line:84% Kateryna, for example, has great prepositional phrases-- 01:10:07.960 --> 01:10:08.590 align:middle line:90% go ahead. 01:10:08.590 --> 01:10:10.132 align:middle line:84% AUDIENCE: Is there a reason why we're 01:10:10.132 --> 01:10:12.280 align:middle line:90% ignoring the intonation angle? 01:10:12.280 --> 01:10:14.657 align:middle line:84% Is this like a competence versus performance situation? 01:10:14.657 --> 01:10:15.490 align:middle line:90% NORVIN RICHARDS: No. 01:10:15.490 --> 01:10:16.330 align:middle line:90% No. 01:10:16.330 --> 01:10:16.900 align:middle line:90% No. 01:10:16.900 --> 01:10:25.180 align:middle line:84% This is more a let's start with imagining 01:10:25.180 --> 01:10:27.860 align:middle line:84% that there's no wind resistance when we're doing physics, 01:10:27.860 --> 01:10:28.360 align:middle line:90% right? 01:10:28.360 --> 01:10:32.230 align:middle line:84% This is a, yeah, that's a really interesting complication, which 01:10:32.230 --> 01:10:36.100 align:middle line:84% we need to study what the heck is going on with intonation. 01:10:36.100 --> 01:10:38.718 align:middle line:84% But we're going to start by developing 01:10:38.718 --> 01:10:41.260 align:middle line:84% a theory of English sentences in which nothing interesting is 01:10:41.260 --> 01:10:42.850 align:middle line:90% happening with intonation. 01:10:42.850 --> 01:10:45.730 align:middle line:84% And then because this is 24.900 we will probably never get out 01:10:45.730 --> 01:10:47.200 align:middle line:90% of that. 01:10:47.200 --> 01:10:50.620 align:middle line:84% But intonation is a fascinating topic 01:10:50.620 --> 01:10:53.260 align:middle line:84% that you then go try to add to your model. 01:10:53.260 --> 01:10:54.490 align:middle line:90% Yeah. 01:10:54.490 --> 01:10:55.240 align:middle line:90% Yes? 01:10:55.240 --> 01:10:58.090 align:middle line:84% AUDIENCE: So to correctly write sentence two 01:10:58.090 --> 01:11:02.170 align:middle line:84% and the sentence below, would you put commas around-- 01:11:02.170 --> 01:11:03.700 align:middle line:90% NORVIN RICHARDS: Oh, yeah. 01:11:03.700 --> 01:11:04.300 align:middle line:90% AUDIENCE: --"on the boat"? 01:11:04.300 --> 01:11:05.425 align:middle line:90% NORVIN RICHARDS: Oh, I see. 01:11:05.425 --> 01:11:07.588 align:middle line:84% So yes, if we-- you mean the string 01:11:07.588 --> 01:11:09.130 align:middle line:84% "I decided on the boat on the plane," 01:11:09.130 --> 01:11:11.620 align:middle line:84% if we wanted it to have reading two? 01:11:11.620 --> 01:11:13.660 align:middle line:84% Yeah, commas are one way that you 01:11:13.660 --> 01:11:17.620 align:middle line:84% indicate the special intonation that makes this possible. 01:11:17.620 --> 01:11:20.860 align:middle line:84% So commas-- no language, actually, 01:11:20.860 --> 01:11:23.920 align:middle line:84% has a very good way to indicate intonation. 01:11:23.920 --> 01:11:26.770 align:middle line:84% It's not something that we're good at spelling, 01:11:26.770 --> 01:11:30.640 align:middle line:84% but punctuation marks are one of the things we do to do that. 01:11:30.640 --> 01:11:31.180 align:middle line:90% Yeah. 01:11:31.180 --> 01:11:32.170 align:middle line:90% Nice point. 01:11:32.170 --> 01:11:32.890 align:middle line:90% Raquel? 01:11:32.890 --> 01:11:34.660 align:middle line:84% AUDIENCE: I was thinking that the rule 01:11:34.660 --> 01:11:37.510 align:middle line:84% that you were saying about the number one 01:11:37.510 --> 01:11:39.310 align:middle line:90% being the preferred computation. 01:11:39.310 --> 01:11:41.490 align:middle line:84% It makes a lot of sense in that the-- 01:11:41.490 --> 01:11:43.240 align:middle line:84% if there was going to be a situation where 01:11:43.240 --> 01:11:45.670 align:middle line:84% a thought two sounded correct, I think 01:11:45.670 --> 01:11:48.130 align:middle line:84% it would have to be something where the adjunct sounds 01:11:48.130 --> 01:11:50.047 align:middle line:84% like it needs to come really close to the verb 01:11:50.047 --> 01:11:51.340 align:middle line:90% decide on a whim. 01:11:51.340 --> 01:11:54.460 align:middle line:84% If you said, "I decided on a whim on the boat," 01:11:54.460 --> 01:11:57.282 align:middle line:84% I feel like I would be more likely to say that made sense 01:11:57.282 --> 01:11:58.990 align:middle line:84% than "I decided on the boat on the plane" 01:11:58.990 --> 01:12:01.300 align:middle line:84% because I need to hear and decide on a whim. 01:12:01.300 --> 01:12:02.440 align:middle line:90% NORVIN RICHARDS: Yeah. 01:12:02.440 --> 01:12:05.680 align:middle line:84% That's a very nice point and a really nice example. 01:12:05.680 --> 01:12:07.750 align:middle line:84% I mean maybe-- so I think the next slide is going 01:12:07.750 --> 01:12:09.370 align:middle line:90% to echo a version of that. 01:12:09.370 --> 01:12:11.560 align:middle line:90% What we're seeing is-- 01:12:11.560 --> 01:12:14.170 align:middle line:84% so here I've got an idealized version of the judgments 01:12:14.170 --> 01:12:16.240 align:middle line:90% that we more or less got. 01:12:16.240 --> 01:12:19.660 align:middle line:84% There are some people for whom you can get either reading. 01:12:19.660 --> 01:12:21.820 align:middle line:84% Interestingly, nobody has the opposite judgment, 01:12:21.820 --> 01:12:27.520 align:middle line:84% that "I decided on the boat on the plane" preferentially means 01:12:27.520 --> 01:12:29.950 align:middle line:84% "I chose the plane while I was on the boat." 01:12:29.950 --> 01:12:35.318 align:middle line:84% And what we're seeing then is that if you have-- 01:12:35.318 --> 01:12:37.360 align:middle line:84% so one difference between arguments and adjuncts, 01:12:37.360 --> 01:12:39.250 align:middle line:84% the reason that I introduced you to them, 01:12:39.250 --> 01:12:41.198 align:middle line:84% is that arguments are picky about which 01:12:41.198 --> 01:12:43.240 align:middle line:84% heads they can combine with and adjuncts are not. 01:12:43.240 --> 01:12:45.490 align:middle line:84% You can do anything with a feather but something 01:12:45.490 --> 01:12:47.593 align:middle line:84% like a direct object, you have to look 01:12:47.593 --> 01:12:49.510 align:middle line:84% at the verb to see whether you have it or not. 01:12:49.510 --> 01:12:52.000 align:middle line:84% And what we're also seeing is that if you have a head that 01:12:52.000 --> 01:12:54.640 align:middle line:84% has both an argument and an adjunct, 01:12:54.640 --> 01:12:57.680 align:middle line:84% and that's the "decide on the boat on the plane" thing, 01:12:57.680 --> 01:13:01.240 align:middle line:84% then the argument is closer to the head. 01:13:01.240 --> 01:13:02.440 align:middle line:90% So far, yeah? 01:13:02.440 --> 01:13:05.470 align:middle line:84% And this connects with Raquel was just saying. 01:13:05.470 --> 01:13:07.398 align:middle line:84% One way to think about it, maybe, is to say, 01:13:07.398 --> 01:13:09.190 align:middle line:84% if you have these two prepositional phrases 01:13:09.190 --> 01:13:10.720 align:middle line:84% and you have this verb and you're 01:13:10.720 --> 01:13:12.760 align:middle line:84% trying to decide, in what order shall I merge 01:13:12.760 --> 01:13:14.230 align:middle line:90% these things with each other? 01:13:14.230 --> 01:13:17.110 align:middle line:84% Apparently the answer is, well, first, 01:13:17.110 --> 01:13:19.730 align:middle line:84% see what the verb would like you to do. 01:13:19.730 --> 01:13:22.900 align:middle line:84% So if the verb needs something, then 01:13:22.900 --> 01:13:26.260 align:middle line:84% satisfy the needs of the verb first and then fool around 01:13:26.260 --> 01:13:26.910 align:middle line:90% with adjuncts. 01:13:26.910 --> 01:13:27.410 align:middle line:90% Yes. 01:13:27.410 --> 01:13:28.960 align:middle line:90% You can do adjuncts later. 01:13:28.960 --> 01:13:31.690 align:middle line:84% And Raquel's making the point that there are various things 01:13:31.690 --> 01:13:34.930 align:middle line:84% in life that verbs could need and we have to develop a theory 01:13:34.930 --> 01:13:37.790 align:middle line:84% of all of them, and today we haven't. 01:13:37.790 --> 01:13:39.240 align:middle line:90% Yep. 01:13:39.240 --> 01:13:41.580 align:middle line:90% Any questions about any of this? 01:13:41.580 --> 01:13:43.837 align:middle line:84% I think, if I remember right, this 01:13:43.837 --> 01:13:45.420 align:middle line:84% might be a good place to pause and see 01:13:45.420 --> 01:13:47.040 align:middle line:90% what I was going to do next. 01:13:47.040 --> 01:13:47.640 align:middle line:90% Yes, no. 01:13:47.640 --> 01:13:49.738 align:middle line:84% So we're about to do another test for arguments 01:13:49.738 --> 01:13:50.280 align:middle line:90% and adjuncts. 01:13:50.280 --> 01:13:53.040 align:middle line:84% And we will do it, but we'll do it next time 01:13:53.040 --> 01:13:56.680 align:middle line:84% where "next time" is not next week, but the week after, 01:13:56.680 --> 01:13:57.180 align:middle line:90% right? 01:13:57.180 --> 01:14:00.480 align:middle line:84% Because you guys have to go break spring. 01:14:00.480 --> 01:14:02.520 align:middle line:84% So have a great spring break, everybody. 01:14:02.520 --> 01:14:05.910 align:middle line:84% Go do something relaxing, and we will see you again 01:14:05.910 --> 01:14:08.330 align:middle line:90% in a couple weeks. 01:14:08.330 --> 01:14:15.000 align:middle line:90%