WEBVTT 00:00:00.000 --> 00:00:02.455 align:middle line:90% [SQUEAKING] 00:00:02.455 --> 00:00:03.428 align:middle line:90% [RUSTLING] 00:00:03.428 --> 00:00:03.928 align:middle line:90% 00:00:03.928 --> 00:00:05.401 align:middle line:90% [CLICKING] 00:00:05.401 --> 00:00:10.595 align:middle line:90% 00:00:10.595 --> 00:00:11.720 align:middle line:90% NORVIN RICHARDS: All right. 00:00:11.720 --> 00:00:15.230 align:middle line:84% So I think this is our last day for semantics. 00:00:15.230 --> 00:00:20.340 align:middle line:84% So we're going to do a grab bag of semantics-y topics today, 00:00:20.340 --> 00:00:23.960 align:middle line:84% and then we will pass on in to other topics starting 00:00:23.960 --> 00:00:26.560 align:middle line:90% next week. 00:00:26.560 --> 00:00:28.720 align:middle line:84% Last time, we ended on a cliffhanger. 00:00:28.720 --> 00:00:31.330 align:middle line:84% So let me remind you where we were. 00:00:31.330 --> 00:00:36.000 align:middle line:84% We said-- here we are with guards and buildings again. 00:00:36.000 --> 00:00:39.250 align:middle line:84% And we said that this sentence was ambiguous. 00:00:39.250 --> 00:00:41.920 align:middle line:84% So "Two guards seemed to me to be standing 00:00:41.920 --> 00:00:45.640 align:middle line:84% in front of every building" has the two familiar readings, 00:00:45.640 --> 00:00:49.270 align:middle line:84% the one where looks to me like every building has two guards, 00:00:49.270 --> 00:00:53.080 align:middle line:84% and the wacky reading where I believe 00:00:53.080 --> 00:00:54.725 align:middle line:84% there are two very large guards who 00:00:54.725 --> 00:00:56.350 align:middle line:84% are guarding all the buildings at once. 00:00:56.350 --> 00:00:57.933 align:middle line:84% You could have both of those readings. 00:00:57.933 --> 00:01:00.700 align:middle line:84% And then we had also convinced ourselves 00:01:00.700 --> 00:01:04.250 align:middle line:84% that that ambiguity goes away in this second sentence. 00:01:04.250 --> 00:01:05.950 align:middle line:84% So "Two guards seemed to themselves 00:01:05.950 --> 00:01:11.705 align:middle line:84% to be standing in front of every building" can only mean-- 00:01:11.705 --> 00:01:13.610 align:middle line:90% this belongs to somebody-- 00:01:13.610 --> 00:01:17.510 align:middle line:84% can only mean that there are two guards who 00:01:17.510 --> 00:01:21.230 align:middle line:84% have a hallucination in which they are very large. 00:01:21.230 --> 00:01:25.070 align:middle line:84% Or there's something else you could imagine it meaning. 00:01:25.070 --> 00:01:27.943 align:middle line:84% It could mean for each building, there 00:01:27.943 --> 00:01:29.360 align:middle line:84% are two guards who think that they 00:01:29.360 --> 00:01:31.310 align:middle line:90% are standing in front of it. 00:01:31.310 --> 00:01:33.650 align:middle line:84% That would, in a way, be more alarming. 00:01:33.650 --> 00:01:36.200 align:middle line:84% There would be more guards who had some kind 00:01:36.200 --> 00:01:38.270 align:middle line:90% of psychological problem. 00:01:38.270 --> 00:01:40.280 align:middle line:84% But they at least wouldn't be believing 00:01:40.280 --> 00:01:43.940 align:middle line:84% that they are impossibly large guards, just that they 00:01:43.940 --> 00:01:45.435 align:middle line:90% believe they're on duty. 00:01:45.435 --> 00:01:47.820 align:middle line:90% [INAUDIBLE] 00:01:47.820 --> 00:01:50.180 align:middle line:84% AUDIENCE: If we arrange the second sentence and say, 00:01:50.180 --> 00:01:51.980 align:middle line:84% "In front of every building, two guards 00:01:51.980 --> 00:01:55.577 align:middle line:84% seem themselves to be standing," is that the same thing? 00:01:55.577 --> 00:01:57.160 align:middle line:84% NORVIN RICHARDS: What do people think? 00:01:57.160 --> 00:01:58.690 align:middle line:84% If I say, "In front of every building, 00:01:58.690 --> 00:02:00.148 align:middle line:84% two guards seem to themselves to be 00:02:00.148 --> 00:02:04.860 align:middle line:84% standing," to the extent that you can say that, 00:02:04.860 --> 00:02:06.660 align:middle line:90% does it mean-- 00:02:06.660 --> 00:02:10.180 align:middle line:90% can it mean there are two-- 00:02:10.180 --> 00:02:12.670 align:middle line:84% for each building, there are twice as many guards 00:02:12.670 --> 00:02:14.258 align:middle line:90% as there are buildings? 00:02:14.258 --> 00:02:16.300 align:middle line:84% Twice as many people who believe themselves to be 00:02:16.300 --> 00:02:18.190 align:middle line:90% guards as there are buildings? 00:02:18.190 --> 00:02:18.760 align:middle line:90% Yes? 00:02:18.760 --> 00:02:23.620 align:middle line:84% AUDIENCE: I think that does stay closer to the two-- 00:02:23.620 --> 00:02:25.570 align:middle line:84% there are two guards for every building. 00:02:25.570 --> 00:02:28.480 align:middle line:84% But I could still interpret it as there 00:02:28.480 --> 00:02:31.003 align:middle line:84% are two very large guards or that the buildings all 00:02:31.003 --> 00:02:31.670 align:middle line:90% face each other. 00:02:31.670 --> 00:02:31.970 align:middle line:90% NORVIN RICHARDS: Yeah. 00:02:31.970 --> 00:02:33.598 align:middle line:84% I agree with you, actually, that it's-- 00:02:33.598 --> 00:02:36.140 align:middle line:84% I think it becomes ambiguous, I think, is what you're saying. 00:02:36.140 --> 00:02:38.390 align:middle line:84% So it gets the meaning that this doesn't have, 00:02:38.390 --> 00:02:40.880 align:middle line:90% but it still has this meaning. 00:02:40.880 --> 00:02:43.020 align:middle line:90% I think you're right about that. 00:02:43.020 --> 00:02:43.520 align:middle line:90% Yeah? 00:02:43.520 --> 00:02:45.560 align:middle line:90% AUDIENCE: I don't know. 00:02:45.560 --> 00:02:48.504 align:middle line:84% Somehow, it does feel to lean towards the other-- 00:02:48.504 --> 00:02:49.962 align:middle line:84% NORVIN RICHARDS: The other meaning. 00:02:49.962 --> 00:02:51.822 align:middle line:84% AUDIENCE: More so than the top one for me. 00:02:51.822 --> 00:02:53.530 align:middle line:84% NORVIN RICHARDS: Oh, I see what you mean. 00:02:53.530 --> 00:02:55.140 align:middle line:84% Yeah, the top one, it's easier to say, 00:02:55.140 --> 00:02:58.140 align:middle line:90% yeah, this is ambiguous. 00:02:58.140 --> 00:02:59.490 align:middle line:90% Yes. 00:02:59.490 --> 00:03:02.790 align:middle line:84% This is kind of reminiscent of when I was first showing you 00:03:02.790 --> 00:03:03.810 align:middle line:90% quantifier scope. 00:03:03.810 --> 00:03:07.620 align:middle line:84% We were talking about sentences like "Everyone in this room 00:03:07.620 --> 00:03:09.360 align:middle line:84% speaks two languages" and "Two languages 00:03:09.360 --> 00:03:11.580 align:middle line:84% are spoken by everyone in this room." 00:03:11.580 --> 00:03:14.653 align:middle line:84% And I think convincing ourselves that whether you have 00:03:14.653 --> 00:03:17.070 align:middle line:84% the active or the passive has some kind of effect on which 00:03:17.070 --> 00:03:20.790 align:middle line:84% reading is easiest to get, and there's something similar going 00:03:20.790 --> 00:03:23.010 align:middle line:84% on here, I think, that there's certainly more 00:03:23.010 --> 00:03:26.880 align:middle line:84% to say than some sentences are ambiguous and others are not. 00:03:26.880 --> 00:03:29.217 align:middle line:84% So some sentences lean in the direction 00:03:29.217 --> 00:03:30.300 align:middle line:90% of one reading or another. 00:03:30.300 --> 00:03:33.420 align:middle line:84% There's a lot of work to do here, which we won't try to do. 00:03:33.420 --> 00:03:37.320 align:middle line:84% This is almost the end of our acquaintance with quantifiers. 00:03:37.320 --> 00:03:39.210 align:middle line:84% But yes, there's lots of interesting topics 00:03:39.210 --> 00:03:40.350 align:middle line:90% to work on here. 00:03:40.350 --> 00:03:41.500 align:middle line:90% Did you have a question? 00:03:41.500 --> 00:03:43.290 align:middle line:84% AUDIENCE: Well, I was just going to say, 00:03:43.290 --> 00:03:46.670 align:middle line:84% rearranging that sentence seems to add extra ambiguity 00:03:46.670 --> 00:03:49.970 align:middle line:84% because now it now gives a sense of whether 00:03:49.970 --> 00:03:52.100 align:middle line:84% or not the guards are standing themselves. 00:03:52.100 --> 00:03:53.285 align:middle line:90% They could be sitting. 00:03:53.285 --> 00:03:55.460 align:middle line:84% They seem to themselves to be standing. 00:03:55.460 --> 00:03:56.780 align:middle line:90% NORVIN RICHARDS: Oh. 00:03:56.780 --> 00:03:57.330 align:middle line:90% Oh. 00:03:57.330 --> 00:03:57.830 align:middle line:90% Oh. 00:03:57.830 --> 00:03:59.630 align:middle line:90% I didn't even think about that. 00:03:59.630 --> 00:04:00.150 align:middle line:90% Right. 00:04:00.150 --> 00:04:00.650 align:middle line:90% Right. 00:04:00.650 --> 00:04:02.240 align:middle line:90% Right. 00:04:02.240 --> 00:04:03.380 align:middle line:90% I was-- yes. 00:04:03.380 --> 00:04:04.160 align:middle line:90% Oh, gosh. 00:04:04.160 --> 00:04:06.140 align:middle line:84% I was thinking about the "seeming"-- 00:04:06.140 --> 00:04:08.570 align:middle line:90% "stand" as just a-- 00:04:08.570 --> 00:04:10.202 align:middle line:90% meaning be, basically. 00:04:10.202 --> 00:04:10.910 align:middle line:90% But you're right. 00:04:10.910 --> 00:04:13.340 align:middle line:84% There's another way that things could seem and not be true. 00:04:13.340 --> 00:04:14.870 align:middle line:84% They could think that they're standing 00:04:14.870 --> 00:04:17.328 align:middle line:84% in front of the buildings, but actually they're lying down. 00:04:17.328 --> 00:04:19.430 align:middle line:84% These are guards with a different kind of problem. 00:04:19.430 --> 00:04:21.649 align:middle line:90% Yeah, good point. 00:04:21.649 --> 00:04:23.440 align:middle line:90% Yeah? 00:04:23.440 --> 00:04:25.210 align:middle line:90% So far so good? 00:04:25.210 --> 00:04:26.980 align:middle line:90% All right. 00:04:26.980 --> 00:04:28.595 align:middle line:84% So ambiguity in the first example, 00:04:28.595 --> 00:04:31.540 align:middle line:84% and ambiguity going away in the second example. 00:04:31.540 --> 00:04:36.100 align:middle line:84% And Raquel thinks, and I think we decided to agree with her, 00:04:36.100 --> 00:04:38.140 align:middle line:84% that you can get the ambiguity back 00:04:38.140 --> 00:04:41.465 align:middle line:84% in that second kind of example by moving "every building." 00:04:41.465 --> 00:04:43.090 align:middle line:84% Let's put Raquel's point aside and just 00:04:43.090 --> 00:04:44.440 align:middle line:90% concentrate on these for a sec. 00:04:44.440 --> 00:04:47.400 align:middle line:90% 00:04:47.400 --> 00:04:50.990 align:middle line:84% Remember way back when, we had the idea 00:04:50.990 --> 00:04:52.970 align:middle line:84% that a sentence like "Two guards seemed 00:04:52.970 --> 00:04:56.130 align:middle line:84% to be standing in front of every building" is ambiguous. 00:04:56.130 --> 00:04:59.060 align:middle line:84% Remember, we decided this ambiguity comes from the fact 00:04:59.060 --> 00:05:01.790 align:middle line:84% that "every building" can, but doesn't 00:05:01.790 --> 00:05:04.850 align:middle line:84% have to, undergo this operation of QR, 00:05:04.850 --> 00:05:09.950 align:middle line:84% quantifier raising, that gets it past two guards. 00:05:09.950 --> 00:05:12.898 align:middle line:84% And we also convinced ourselves that this operation 00:05:12.898 --> 00:05:14.690 align:middle line:84% of quantifier raising has some restrictions 00:05:14.690 --> 00:05:17.123 align:middle line:90% on how far it can go. 00:05:17.123 --> 00:05:19.040 align:middle line:84% The evidence for that had to do with sentences 00:05:19.040 --> 00:05:20.750 align:middle line:90% like the last one on this slide. 00:05:20.750 --> 00:05:22.340 align:middle line:84% "Two guards think that I am standing 00:05:22.340 --> 00:05:27.020 align:middle line:84% in front of every building" isn't ambiguous. 00:05:27.020 --> 00:05:29.780 align:middle line:84% Doesn't mean-- so it can only mean 00:05:29.780 --> 00:05:35.530 align:middle line:84% there are two guards, two particular guards, who 00:05:35.530 --> 00:05:39.220 align:middle line:84% are having real problems who seem to see me 00:05:39.220 --> 00:05:40.780 align:middle line:90% in front of every building. 00:05:40.780 --> 00:05:43.120 align:middle line:84% It doesn't mean there are two guards who 00:05:43.120 --> 00:05:44.553 align:middle line:84% think I'm in front of building 32, 00:05:44.553 --> 00:05:45.970 align:middle line:84% and there are two other guards who 00:05:45.970 --> 00:05:47.410 align:middle line:84% think I'm in front of the Student Center, 00:05:47.410 --> 00:05:48.400 align:middle line:84% and there are two other guards who 00:05:48.400 --> 00:05:50.025 align:middle line:84% think that I'm in front of the library. 00:05:50.025 --> 00:05:51.580 align:middle line:90% It doesn't mean that. 00:05:51.580 --> 00:05:54.370 align:middle line:84% So it only has the reading where "every building" hasn't 00:05:54.370 --> 00:05:56.410 align:middle line:90% undergone QR past "two guards." 00:05:56.410 --> 00:05:58.270 align:middle line:84% It only has the meaning where you 00:05:58.270 --> 00:05:59.690 align:middle line:90% interpret "two guards" first. 00:05:59.690 --> 00:06:02.050 align:middle line:84% It's like, there are two guards who have this property. 00:06:02.050 --> 00:06:04.383 align:middle line:84% They think that I'm standing in front of every building. 00:06:04.383 --> 00:06:06.460 align:middle line:84% It doesn't mean it's true of every building. 00:06:06.460 --> 00:06:09.490 align:middle line:84% The set of buildings and the set of things that two guards think 00:06:09.490 --> 00:06:14.470 align:middle line:84% that I am standing in front of has the following intersection. 00:06:14.470 --> 00:06:16.255 align:middle line:90% Do people see that? 00:06:16.255 --> 00:06:18.130 align:middle line:84% So there's no ambiguity in that last example. 00:06:18.130 --> 00:06:20.440 align:middle line:84% And our story about why there's no ambiguity 00:06:20.440 --> 00:06:23.440 align:middle line:84% in that last example, this description of that fact 00:06:23.440 --> 00:06:25.690 align:middle line:84% was to say, OK, we're learning something about QR. 00:06:25.690 --> 00:06:28.450 align:middle line:84% QR can't-- when you have two guards that are standing 00:06:28.450 --> 00:06:31.390 align:middle line:84% in front of every building, "two guards" and "every building" 00:06:31.390 --> 00:06:33.650 align:middle line:84% are close enough that QR can reorder them, 00:06:33.650 --> 00:06:34.720 align:middle line:90% and you get ambiguity. 00:06:34.720 --> 00:06:36.310 align:middle line:84% But in this one, "Two guards think 00:06:36.310 --> 00:06:38.227 align:middle line:84% that I'm standing in front of every building," 00:06:38.227 --> 00:06:39.970 align:middle line:84% "every building" can't QR that far. 00:06:39.970 --> 00:06:41.570 align:middle line:84% And then that gets us into question. 00:06:41.570 --> 00:06:43.120 align:middle line:90% So how far can QR go? 00:06:43.120 --> 00:06:45.370 align:middle line:90% How does it work? 00:06:45.370 --> 00:06:46.990 align:middle line:90% We got started on that question. 00:06:46.990 --> 00:06:48.160 align:middle line:90% Do people remember all this? 00:06:48.160 --> 00:06:49.360 align:middle line:90% This is all review. 00:06:49.360 --> 00:06:51.580 align:middle line:90% This might make sense. 00:06:51.580 --> 00:06:53.800 align:middle line:84% Did you have a question, Katrina? 00:06:53.800 --> 00:06:54.730 align:middle line:90% AUDIENCE: Yeah. 00:06:54.730 --> 00:06:57.220 align:middle line:84% What does it mean that "two guards" can reconstruct back 00:06:57.220 --> 00:06:58.310 align:middle line:90% into the embedded clause? 00:06:58.310 --> 00:06:59.500 align:middle line:90% NORVIN RICHARDS: Oh, OK. 00:06:59.500 --> 00:07:01.550 align:middle line:84% I'm sorry, I'm doing this slide out of order. 00:07:01.550 --> 00:07:03.348 align:middle line:84% So I'm talking about the last sentence. 00:07:03.348 --> 00:07:05.140 align:middle line:84% In the last sentence, there's no ambiguity, 00:07:05.140 --> 00:07:07.557 align:middle line:84% and we convinced ourselves that's because "every building" 00:07:07.557 --> 00:07:09.160 align:middle line:84% can't get out of that embedded clause. 00:07:09.160 --> 00:07:11.590 align:middle line:84% So question, what counts as a clause? 00:07:11.590 --> 00:07:13.780 align:middle line:90% How far can QR go? 00:07:13.780 --> 00:07:15.940 align:middle line:84% The fact that there's ambiguity in "Two guards 00:07:15.940 --> 00:07:18.580 align:middle line:84% seem to be standing in front of every building," we said, 00:07:18.580 --> 00:07:20.622 align:middle line:84% we could have handled that in two different ways. 00:07:20.622 --> 00:07:23.560 align:middle line:84% Here, let me write that on the board. 00:07:23.560 --> 00:07:24.910 align:middle line:90% For some reason, there are-- 00:07:24.910 --> 00:07:25.782 align:middle line:90% here we go. 00:07:25.782 --> 00:07:28.240 align:middle line:84% There are many erasers on the board, but there is no chalk. 00:07:28.240 --> 00:07:30.910 align:middle line:84% I think someone is trying to send me a message. 00:07:30.910 --> 00:07:41.650 align:middle line:84% So "Two guards seem to be standing 00:07:41.650 --> 00:07:45.295 align:middle line:90% in front of every building." 00:07:45.295 --> 00:07:49.640 align:middle line:90% 00:07:49.640 --> 00:07:53.650 align:middle line:84% So recall that for independent reasons, 00:07:53.650 --> 00:07:55.540 align:middle line:90% we thought that "two guards"-- 00:07:55.540 --> 00:07:57.040 align:middle line:84% in a sentence like this, "Two guards 00:07:57.040 --> 00:07:59.582 align:middle line:84% seem to be standing in front of every building," "two guards" 00:07:59.582 --> 00:08:01.870 align:middle line:84% is here, but it starts down here and moves. 00:08:01.870 --> 00:08:05.320 align:middle line:84% We have this NP movement that took things 00:08:05.320 --> 00:08:07.240 align:middle line:84% from embedded clauses and moved them 00:08:07.240 --> 00:08:10.480 align:middle line:84% into the subject of things like "seem" 00:08:10.480 --> 00:08:13.640 align:middle line:84% in order to satisfy the EPP was the story. 00:08:13.640 --> 00:08:16.400 align:middle line:84% So we have this operation and movement. 00:08:16.400 --> 00:08:18.940 align:middle line:84% So now there's this open question. 00:08:18.940 --> 00:08:23.920 align:middle line:84% We want QR to be able to get "every building" past "two 00:08:23.920 --> 00:08:25.090 align:middle line:90% guards." 00:08:25.090 --> 00:08:28.090 align:middle line:84% But "two guards" is in two places. 00:08:28.090 --> 00:08:29.305 align:middle line:90% It's in the lower position. 00:08:29.305 --> 00:08:30.680 align:middle line:84% It's also in the higher position. 00:08:30.680 --> 00:08:34.240 align:middle line:84% So we could say QR can go from here all the way 00:08:34.240 --> 00:08:35.919 align:middle line:90% to the beginning. 00:08:35.919 --> 00:08:38.830 align:middle line:84% That's a way of saying that embedded clause does not 00:08:38.830 --> 00:08:40.840 align:middle line:90% count as a clause. 00:08:40.840 --> 00:08:44.230 align:middle line:84% QR can get out of that kind of clause, that infinitival kind 00:08:44.230 --> 00:08:45.018 align:middle line:90% of clause. 00:08:45.018 --> 00:08:46.810 align:middle line:84% But another kind of thing that we could say 00:08:46.810 --> 00:08:49.600 align:middle line:84% would be no, the reason this is ambiguous 00:08:49.600 --> 00:08:51.970 align:middle line:84% is that "every building" can get this far 00:08:51.970 --> 00:08:54.770 align:middle line:84% to the edge of the embedded clause. 00:08:54.770 --> 00:08:59.770 align:middle line:84% And so it can't, in fact, get this far. 00:08:59.770 --> 00:09:01.570 align:middle line:84% Can't get out of that embedded clause. 00:09:01.570 --> 00:09:09.760 align:middle line:84% Basically, this involves saying QR can't escape TP. 00:09:09.760 --> 00:09:12.220 align:middle line:84% So it can't escape any kind of TP. 00:09:12.220 --> 00:09:15.610 align:middle line:84% Not a finite clause like the one in the last example 00:09:15.610 --> 00:09:17.110 align:middle line:84% here, "Two guards think I'm standing 00:09:17.110 --> 00:09:21.580 align:middle line:84% in front of every building," and also not an infinitival clause. 00:09:21.580 --> 00:09:24.220 align:middle line:84% And we saw before some reasons to think that that's right, 00:09:24.220 --> 00:09:26.185 align:middle line:84% that the only reason this is ambiguous, 00:09:26.185 --> 00:09:28.060 align:middle line:84% it's not because "every building" can get out 00:09:28.060 --> 00:09:28.808 align:middle line:90% of an infinitive. 00:09:28.808 --> 00:09:30.100 align:middle line:90% "Every building" is stuck here. 00:09:30.100 --> 00:09:31.480 align:middle line:84% It can't get any higher than this. 00:09:31.480 --> 00:09:33.970 align:middle line:84% And the reason that this is ambiguous is that "two guards," 00:09:33.970 --> 00:09:37.180 align:middle line:84% yes, it's pronounced here, but it can be interpreted here 00:09:37.180 --> 00:09:39.370 align:middle line:90% via reconstruction. 00:09:39.370 --> 00:09:41.420 align:middle line:84% Which is just a name for this mysterious fact. 00:09:41.420 --> 00:09:43.630 align:middle line:84% You can interpret things lower if they 00:09:43.630 --> 00:09:45.100 align:middle line:90% moved from lower positions. 00:09:45.100 --> 00:09:47.650 align:middle line:84% We saw some evidence for that idea before. 00:09:47.650 --> 00:09:50.230 align:middle line:84% I'm about to show you another bit of evidence for that idea. 00:09:50.230 --> 00:09:53.260 align:middle line:90% But is this much clear? 00:09:53.260 --> 00:09:54.850 align:middle line:90% It's late in the semester. 00:09:54.850 --> 00:09:57.950 align:middle line:84% Probably almost nothing is clear anymore. 00:09:57.950 --> 00:09:59.440 align:middle line:90% It's spring outside. 00:09:59.440 --> 00:10:02.530 align:middle line:84% Yeah, it's very hard to make things clear. 00:10:02.530 --> 00:10:04.810 align:middle line:84% Do people want to ask questions about this? 00:10:04.810 --> 00:10:06.730 align:middle line:90% OK, all right. 00:10:06.730 --> 00:10:09.200 align:middle line:84% So here's a new bit of ambiguity, 00:10:09.200 --> 00:10:11.190 align:middle line:84% a new bit of evidence that this is right, 00:10:11.190 --> 00:10:12.820 align:middle line:84% that this is the right way of handling 00:10:12.820 --> 00:10:14.980 align:middle line:84% the ambiguity for the first example in this slide, 00:10:14.980 --> 00:10:17.480 align:middle line:84% "Two guards seem to be standing in front of every building." 00:10:17.480 --> 00:10:19.810 align:middle line:84% So we'll say, the reason that's ambiguous 00:10:19.810 --> 00:10:23.752 align:middle line:84% is that QR, first of all, can't get out of TP. 00:10:23.752 --> 00:10:25.960 align:middle line:84% It can't get out of that embedded infinitival clause. 00:10:25.960 --> 00:10:28.100 align:middle line:90% The earliest it can get is here. 00:10:28.100 --> 00:10:30.200 align:middle line:84% So "to be standing in front of every building." 00:10:30.200 --> 00:10:31.420 align:middle line:84% And the reason you have ambiguity 00:10:31.420 --> 00:10:34.003 align:middle line:84% there is that "two guards" has the option of being interpreted 00:10:34.003 --> 00:10:36.978 align:middle line:84% down here, this mysterious option that we've been seeing 00:10:36.978 --> 00:10:39.520 align:middle line:84% for movement, that you have the option of interpreting things 00:10:39.520 --> 00:10:41.920 align:middle line:84% as though they were lower than they actually are, 00:10:41.920 --> 00:10:46.620 align:middle line:84% as though they are still where they used to be. 00:10:46.620 --> 00:10:48.910 align:middle line:90% Here's the bit of evidence. 00:10:48.910 --> 00:10:50.620 align:middle line:84% "Two guards seem to themselves to be 00:10:50.620 --> 00:10:52.162 align:middle line:84% standing in front of every building," 00:10:52.162 --> 00:10:55.240 align:middle line:90% we decided that's not ambiguous. 00:10:55.240 --> 00:10:59.270 align:middle line:84% "Every building" can't take scope above "two guards." 00:10:59.270 --> 00:11:04.250 align:middle line:84% Well, let's think about what would happen. 00:11:04.250 --> 00:11:06.410 align:middle line:84% I'm going to add "to themselves" here. 00:11:06.410 --> 00:11:09.890 align:middle line:90% 00:11:09.890 --> 00:11:11.600 align:middle line:84% We're fooling around with the possibility 00:11:11.600 --> 00:11:14.260 align:middle line:84% that the ambiguity in "Two guards seemed 00:11:14.260 --> 00:11:16.010 align:middle line:84% to be standing in front of every building" 00:11:16.010 --> 00:11:17.450 align:middle line:84% is there because "two guards" has 00:11:17.450 --> 00:11:19.592 align:middle line:84% the option of reconstructing into the lower clause, 00:11:19.592 --> 00:11:21.050 align:middle line:84% has the option of being interpreted 00:11:21.050 --> 00:11:23.270 align:middle line:84% as though it was still down here. 00:11:23.270 --> 00:11:26.580 align:middle line:84% But that's why that sentence is ambiguous. 00:11:26.580 --> 00:11:29.720 align:middle line:84% So why isn't this sentence ambiguous? 00:11:29.720 --> 00:11:32.770 align:middle line:84% Well, what we want is for something 00:11:32.770 --> 00:11:35.270 align:middle line:84% to prevent the "two guards" from reconstructing down to here 00:11:35.270 --> 00:11:38.390 align:middle line:84% if the "two guards" had to be up here, if they couldn't 00:11:38.390 --> 00:11:40.790 align:middle line:84% be interpreted in this lower position. 00:11:40.790 --> 00:11:43.130 align:middle line:84% Well, then QR wouldn't be able to get past them 00:11:43.130 --> 00:11:47.127 align:middle line:84% because the hypothesis is that QR can only get this far. 00:11:47.127 --> 00:11:48.960 align:middle line:84% Is there something that will stop the guards 00:11:48.960 --> 00:11:51.290 align:middle line:90% from reconstructing? 00:11:51.290 --> 00:11:53.090 align:middle line:84% Why shouldn't "two guards"-- suppose 00:11:53.090 --> 00:11:54.560 align:middle line:84% we got rid of "two guards" up here. 00:11:54.560 --> 00:11:59.150 align:middle line:84% So we would have "Seemed to themselves, two guards, 00:11:59.150 --> 00:12:01.872 align:middle line:84% to be standing in front of every building." 00:12:01.872 --> 00:12:04.205 align:middle line:84% Would anything go wrong with the interpretation of that? 00:12:04.205 --> 00:12:06.900 align:middle line:90% 00:12:06.900 --> 00:12:10.070 align:middle line:84% Recall that what makes the sentence unambiguous is 00:12:10.070 --> 00:12:11.510 align:middle line:90% putting in "to themselves." 00:12:11.510 --> 00:12:14.565 align:middle line:84% If "to themselves" is not here, if it's just seem, 00:12:14.565 --> 00:12:16.190 align:middle line:84% well then "two guards" can reconstruct. 00:12:16.190 --> 00:12:18.273 align:middle line:84% But when we add "to themselves," it can't anymore. 00:12:18.273 --> 00:12:21.630 align:middle line:90% It has to stay up high. 00:12:21.630 --> 00:12:24.270 align:middle line:90% What's "themselves"? 00:12:24.270 --> 00:12:25.440 align:middle line:90% (What are "themselves"? 00:12:25.440 --> 00:12:26.640 align:middle line:90% What am "themselves"? 00:12:26.640 --> 00:12:27.948 align:middle line:90% What??) 00:12:27.948 --> 00:12:29.280 align:middle line:90% AUDIENCE: Anaphore? 00:12:29.280 --> 00:12:31.290 align:middle line:84% NORVIN RICHARDS: It's an anaphore, right? 00:12:31.290 --> 00:12:32.670 align:middle line:90% It's a reflexive. 00:12:32.670 --> 00:12:33.660 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:12:33.660 --> 00:12:36.035 align:middle line:84% NORVIN RICHARDS: Yeah, that's what you were about to say. 00:12:36.035 --> 00:12:37.433 align:middle line:90% It's a reflexive. 00:12:37.433 --> 00:12:38.725 align:middle line:90% Somebody just had a great idea. 00:12:38.725 --> 00:12:41.660 align:middle line:90% 00:12:41.660 --> 00:12:42.550 align:middle line:90% It's a reflexive. 00:12:42.550 --> 00:12:45.550 align:middle line:84% Reflexives are subject to condition A, right? 00:12:45.550 --> 00:12:46.720 align:middle line:90% What does condition A say? 00:12:46.720 --> 00:12:48.175 align:middle line:90% What do reflexives need in life? 00:12:48.175 --> 00:12:51.729 align:middle line:90% 00:12:51.729 --> 00:12:53.130 align:middle line:90% AUDIENCE: They have to be bound. 00:12:53.130 --> 00:12:54.570 align:middle line:84% NORVIN RICHARDS: They have to be bound, exactly. 00:12:54.570 --> 00:12:55.980 align:middle line:84% They need something to bind them. 00:12:55.980 --> 00:12:57.647 align:middle line:84% They need something that c-commands them 00:12:57.647 --> 00:12:59.700 align:middle line:90% and co-refers with them. 00:12:59.700 --> 00:13:03.280 align:middle line:84% That's why you can say, "The guards like themselves," 00:13:03.280 --> 00:13:06.000 align:middle line:84% but not Themselves like the guards." 00:13:06.000 --> 00:13:07.500 align:middle line:84% "Themselves" needs to be c-commanded 00:13:07.500 --> 00:13:10.430 align:middle line:84% by "guards," or some other noun that refers to them. 00:13:10.430 --> 00:13:12.960 align:middle line:90% 00:13:12.960 --> 00:13:15.540 align:middle line:84% So what would go wrong if you reconstructed "the guards" down 00:13:15.540 --> 00:13:17.460 align:middle line:84% here if you have "themselves" here 00:13:17.460 --> 00:13:20.410 align:middle line:84% is that, yeah, "guards" would be able to be scope ambiguous 00:13:20.410 --> 00:13:21.660 align:middle line:90% with respect to the buildings. 00:13:21.660 --> 00:13:26.830 align:middle line:84% But "themselves" would not have a binder anymore. 00:13:26.830 --> 00:13:31.800 align:middle line:84% What we're learning-- so when I was first showing you 00:13:31.800 --> 00:13:36.540 align:middle line:84% some examples of reconstruction, I think I said something like, 00:13:36.540 --> 00:13:41.170 align:middle line:84% it's as though anaphores get to look back on the past and say, 00:13:41.170 --> 00:13:43.952 align:middle line:84% well, I'm not bound anymore, but I used to be. 00:13:43.952 --> 00:13:44.910 align:middle line:90% And that's good enough. 00:13:44.910 --> 00:13:46.393 align:middle line:90% Something like that. 00:13:46.393 --> 00:13:48.060 align:middle line:84% This is showing that that's not actually 00:13:48.060 --> 00:13:50.143 align:middle line:84% quite the right way to think about reconstruction. 00:13:50.143 --> 00:13:53.440 align:middle line:84% It's more like if you move something, 00:13:53.440 --> 00:13:56.580 align:middle line:84% it exists in various positions in the clause. 00:13:56.580 --> 00:14:00.420 align:middle line:84% But you have to pick one for interpretation. 00:14:00.420 --> 00:14:03.000 align:middle line:84% For things like quantifier scope and binding, 00:14:03.000 --> 00:14:06.990 align:middle line:84% you've got all of these places where it is simultaneously. 00:14:06.990 --> 00:14:10.860 align:middle line:84% But the waveform has to collapse in order 00:14:10.860 --> 00:14:15.030 align:middle line:84% to give it a scope and an interpretation for binding. 00:14:15.030 --> 00:14:19.380 align:middle line:84% You have to pick one of the positions for it to be in. 00:14:19.380 --> 00:14:21.510 align:middle line:84% You run the experiment, you detect 00:14:21.510 --> 00:14:23.430 align:middle line:84% the electron or whatever it is, and then 00:14:23.430 --> 00:14:26.200 align:middle line:90% you know where the guards are. 00:14:26.200 --> 00:14:27.630 align:middle line:84% And once where they are, they have 00:14:27.630 --> 00:14:32.550 align:middle line:84% to be in a position where they behave as being just 00:14:32.550 --> 00:14:35.805 align:middle line:84% in that position both for anaphore binding and for scope. 00:14:35.805 --> 00:14:37.680 align:middle line:84% That's what these data seem to be telling us. 00:14:37.680 --> 00:14:39.243 align:middle line:84% So "Two guards seem to be standing 00:14:39.243 --> 00:14:40.410 align:middle line:90% in front of every building." 00:14:40.410 --> 00:14:42.995 align:middle line:84% "Two guards" is in two places simultaneously. 00:14:42.995 --> 00:14:45.120 align:middle line:84% And you get to choose, is it in the higher position 00:14:45.120 --> 00:14:46.485 align:middle line:90% or the lower position? 00:14:46.485 --> 00:14:48.360 align:middle line:84% "Two guards seem to themselves to be standing 00:14:48.360 --> 00:14:50.277 align:middle line:84% in front of every building" forces you to pick 00:14:50.277 --> 00:14:51.120 align:middle line:90% a particular choice. 00:14:51.120 --> 00:14:53.670 align:middle line:84% You have to keep "the guards" up high so that "themselves" 00:14:53.670 --> 00:14:54.660 align:middle line:90% will have a binder. 00:14:54.660 --> 00:14:57.810 align:middle line:84% And by choosing that, you also make 00:14:57.810 --> 00:15:00.222 align:middle line:84% a choice about how quantifier scope is going to work. 00:15:00.222 --> 00:15:01.680 align:middle line:84% You make it so that the quantifiers 00:15:01.680 --> 00:15:03.360 align:middle line:84% can't be ambiguous anymore because the guards are 00:15:03.360 --> 00:15:03.990 align:middle line:90% too high. 00:15:03.990 --> 00:15:05.670 align:middle line:84% The buildings cannot get to them. 00:15:05.670 --> 00:15:08.220 align:middle line:90% QR doesn't go that far. 00:15:08.220 --> 00:15:10.970 align:middle line:84% That's the way quantifier scope seems to work. 00:15:10.970 --> 00:15:14.740 align:middle line:84% It's the way reconstruction seems to work. 00:15:14.740 --> 00:15:17.850 align:middle line:84% So the moral of all this is reconstruction 00:15:17.850 --> 00:15:20.100 align:middle line:84% involves picking a particular position 00:15:20.100 --> 00:15:22.740 align:middle line:90% to interpret a moved phrase in. 00:15:22.740 --> 00:15:25.290 align:middle line:84% I gave you this idea or this metaphor 00:15:25.290 --> 00:15:27.450 align:middle line:84% of you've got this thing, which is moving, 00:15:27.450 --> 00:15:31.182 align:middle line:84% and you get to rewind the tape and decide-- 00:15:31.182 --> 00:15:32.640 align:middle line:84% look at it in a different position, 00:15:32.640 --> 00:15:35.100 align:middle line:84% not the position where it's actually being pronounced. 00:15:35.100 --> 00:15:37.780 align:middle line:84% And you can think about it that way if you want. 00:15:37.780 --> 00:15:41.340 align:middle line:84% But the point is you have to pick one particular position 00:15:41.340 --> 00:15:43.440 align:middle line:90% as the position where it gets-- 00:15:43.440 --> 00:15:47.850 align:middle line:84% you don't get to say, OK, it'll be here for anaphore binding, 00:15:47.850 --> 00:15:51.183 align:middle line:84% but for quantifier scope, it will be here. 00:15:51.183 --> 00:15:52.350 align:middle line:90% That's not the way it works. 00:15:52.350 --> 00:15:54.690 align:middle line:84% You pick one spot for it to be for all 00:15:54.690 --> 00:15:58.080 align:middle line:90% of its interpretive properties. 00:15:58.080 --> 00:16:00.630 align:middle line:84% This way of-- this fact about reconstruction, 00:16:00.630 --> 00:16:02.730 align:middle line:84% just the existence of reconstruction 00:16:02.730 --> 00:16:04.590 align:middle line:84% and this fact about how it works, 00:16:04.590 --> 00:16:07.980 align:middle line:84% is sometimes taken as a bit of evidence 00:16:07.980 --> 00:16:11.040 align:middle line:84% for a particular way of talking about movement, what's called 00:16:11.040 --> 00:16:12.360 align:middle line:90% the copy theory of movement. 00:16:12.360 --> 00:16:14.040 align:middle line:84% And so I just want to show you that 00:16:14.040 --> 00:16:16.680 align:middle line:84% and then we're going to leave this whole area behind 00:16:16.680 --> 00:16:18.180 align:middle line:90% for a little bit. 00:16:18.180 --> 00:16:19.450 align:middle line:90% The copy theory of movement. 00:16:19.450 --> 00:16:23.610 align:middle line:84% So when we started syntax, cast your mind 00:16:23.610 --> 00:16:25.678 align:middle line:84% back to when we started syntax, when you 00:16:25.678 --> 00:16:26.970 align:middle line:90% were younger and more carefree. 00:16:26.970 --> 00:16:29.520 align:middle line:90% 00:16:29.520 --> 00:16:31.440 align:middle line:84% When we started syntax, I started 00:16:31.440 --> 00:16:34.800 align:middle line:84% by saying something like, syntax is 00:16:34.800 --> 00:16:36.960 align:middle line:84% going to be a little bit like morphology. 00:16:36.960 --> 00:16:39.930 align:middle line:84% So cast your mind back to morphology. 00:16:39.930 --> 00:16:42.030 align:middle line:84% We said we're going to create these tree 00:16:42.030 --> 00:16:45.775 align:middle line:84% structures via repeated application of merge. 00:16:45.775 --> 00:16:47.400 align:middle line:84% So we're going to take things and we're 00:16:47.400 --> 00:16:50.370 align:middle line:84% going to put them together and make larger things. 00:16:50.370 --> 00:16:53.130 align:middle line:84% And Merge is going to be recursive in the sense 00:16:53.130 --> 00:16:55.260 align:middle line:84% that it gets to take things that it created 00:16:55.260 --> 00:16:57.160 align:middle line:84% and merge them with other things. 00:16:57.160 --> 00:16:58.920 align:middle line:84% So you can merge things that you take out 00:16:58.920 --> 00:17:01.496 align:middle line:84% of the lexicon, morphemes, or words, or whatever, 00:17:01.496 --> 00:17:04.079 align:middle line:84% and you can merge those things together to form larger things. 00:17:04.079 --> 00:17:07.950 align:middle line:84% But you can also merge the phrases you've created together 00:17:07.950 --> 00:17:08.760 align:middle line:90% with other phrases. 00:17:08.760 --> 00:17:11.010 align:middle line:84% So for example, suppose we want to construct 00:17:11.010 --> 00:17:13.944 align:middle line:84% the embedded clause of "I don't know what Mary will eat," 00:17:13.944 --> 00:17:15.569 align:middle line:84% where I'm using an embedded clause just 00:17:15.569 --> 00:17:18.640 align:middle line:84% so we won't have to think about what the auxiliary does. 00:17:18.640 --> 00:17:24.000 align:middle line:84% So the idea was you're going to do the embedded clause 00:17:24.000 --> 00:17:26.819 align:middle line:90% by merging "eat" with "what." 00:17:26.819 --> 00:17:29.160 align:middle line:84% Remember, we have the idea that forming 00:17:29.160 --> 00:17:33.180 align:middle line:84% this kind of embedded clause involves moving "what." 00:17:33.180 --> 00:17:35.790 align:middle line:84% And so we're going to start "what" as the object of "eat." 00:17:35.790 --> 00:17:38.070 align:middle line:84% So we'll take these two things from the lexicon, "eat" 00:17:38.070 --> 00:17:40.080 align:middle line:84% and "what," and we'll merge them together 00:17:40.080 --> 00:17:42.960 align:middle line:84% to make this bigger thing, this verb phrase. 00:17:42.960 --> 00:17:45.900 align:middle line:84% And then we'll take that thing, that verb phrase that we 00:17:45.900 --> 00:17:48.210 align:middle line:84% created, and we'll merge that together with a new thing 00:17:48.210 --> 00:17:49.950 align:middle line:84% that we'll get from the lexicon, "will." 00:17:49.950 --> 00:17:51.300 align:middle line:90% Yeah, so we'll take "will." 00:17:51.300 --> 00:17:53.290 align:middle line:84% We'll put that together with the verb phrase. 00:17:53.290 --> 00:17:56.350 align:middle line:84% Now we have an even larger thing and we have "will eat what." 00:17:56.350 --> 00:17:58.352 align:middle line:84% And then we take that thing that we've made, 00:17:58.352 --> 00:18:00.060 align:middle line:84% and we merge it together with a new thing 00:18:00.060 --> 00:18:02.160 align:middle line:84% that we get out of the lexicon, "Mary." 00:18:02.160 --> 00:18:05.580 align:middle line:84% So we take "Mary," and merge that together with T-bar. 00:18:05.580 --> 00:18:09.640 align:middle line:84% As a comparatively simple example, if it were, 00:18:09.640 --> 00:18:12.240 align:middle line:84% "I don't know what the students will eat," 00:18:12.240 --> 00:18:14.160 align:middle line:84% then I'd have a bit where it had to create 00:18:14.160 --> 00:18:16.215 align:middle line:84% "the students" by merging "the" with "students." 00:18:16.215 --> 00:18:17.730 align:middle line:90% I have a noun phrase. 00:18:17.730 --> 00:18:22.230 align:middle line:84% I merge my noun phrase with my T-bar, and now I've got my TP. 00:18:22.230 --> 00:18:23.100 align:middle line:90% So far so good? 00:18:23.100 --> 00:18:25.855 align:middle line:84% This is-- I'm just building a tree. 00:18:25.855 --> 00:18:28.230 align:middle line:84% One more step, I'll grab one more thing from the lexicon. 00:18:28.230 --> 00:18:32.550 align:middle line:84% We'll merge a C with what we've created so far, a C and a TP. 00:18:32.550 --> 00:18:34.950 align:middle line:84% And now it's time to do wh-movement. 00:18:34.950 --> 00:18:38.880 align:middle line:84% And the idea behind the copy theory of movement is look, 00:18:38.880 --> 00:18:41.220 align:middle line:84% this is just another case of Merge. 00:18:41.220 --> 00:18:43.650 align:middle line:84% So all of our cases of Merge, we've 00:18:43.650 --> 00:18:48.090 align:middle line:84% always had the option of merging either something we've created 00:18:48.090 --> 00:18:51.267 align:middle line:84% via Merge, we've merged the VP with something, 00:18:51.267 --> 00:18:53.850 align:middle line:84% merged the T-bar with something, merged the TP with something, 00:18:53.850 --> 00:18:55.860 align:middle line:84% or grabbing something new from the lexicon 00:18:55.860 --> 00:18:57.230 align:middle line:90% and putting that in. 00:18:57.230 --> 00:18:59.880 align:middle line:84% We've always had that option right along. 00:18:59.880 --> 00:19:03.240 align:middle line:84% And the idea is this thing that we're calling movement, 00:19:03.240 --> 00:19:05.250 align:middle line:84% it's just another example of Merge. 00:19:05.250 --> 00:19:08.040 align:middle line:84% We'll take this C-bar that we've created 00:19:08.040 --> 00:19:10.200 align:middle line:84% and we'll merge it with something that's already 00:19:10.200 --> 00:19:12.870 align:middle line:90% in the tree, with "what"? 00:19:12.870 --> 00:19:16.180 align:middle line:84% So we've got-- now "what" is in two places. 00:19:16.180 --> 00:19:18.330 align:middle line:84% So it's in the place where we merged it first 00:19:18.330 --> 00:19:20.375 align:middle line:84% and it's in this new place that's up at the top. 00:19:20.375 --> 00:19:24.630 align:middle line:84% And so we end up with "What Mary will eat what," 00:19:24.630 --> 00:19:27.180 align:middle line:90% which raises certain questions. 00:19:27.180 --> 00:19:30.870 align:middle line:84% Like why don't we say, "What Mary will eat what?" 00:19:30.870 --> 00:19:33.780 align:middle line:84% So we need a theory of what happens when you merge 00:19:33.780 --> 00:19:37.050 align:middle line:90% a single thing in two places. 00:19:37.050 --> 00:19:39.210 align:middle line:84% But the result is that "what" is now 00:19:39.210 --> 00:19:43.230 align:middle line:84% in more than one place in the tree. 00:19:43.230 --> 00:19:45.420 align:middle line:90% And that's all that Merge-- 00:19:45.420 --> 00:19:47.070 align:middle line:90% that's all that movement is. 00:19:47.070 --> 00:19:48.570 align:middle line:84% So I've been showing you trees where 00:19:48.570 --> 00:19:51.820 align:middle line:84% we created trees via Merge and then there were arrows. 00:19:51.820 --> 00:19:53.400 align:middle line:84% So you took things out of the places 00:19:53.400 --> 00:19:55.400 align:middle line:84% where they were, and you put them in new places, 00:19:55.400 --> 00:19:57.720 align:middle line:84% and they weren't in the old places anymore. 00:19:57.720 --> 00:19:59.610 align:middle line:84% And now all this talk about reconstruction 00:19:59.610 --> 00:20:01.710 align:middle line:84% makes it look like that was misleading. 00:20:01.710 --> 00:20:04.440 align:middle line:84% There's a sense in which thing-- when you move things, 00:20:04.440 --> 00:20:09.015 align:middle line:84% they're in all of the places where they ever were. 00:20:09.015 --> 00:20:10.890 align:middle line:84% For interpretation, it sure looks like that's 00:20:10.890 --> 00:20:12.090 align:middle line:90% the way to think about it. 00:20:12.090 --> 00:20:15.210 align:middle line:84% You can interpret them as though they were in any of the places 00:20:15.210 --> 00:20:17.520 align:middle line:90% that they've ever been. 00:20:17.520 --> 00:20:19.290 align:middle line:84% And the copy theory of movement is meant 00:20:19.290 --> 00:20:21.195 align:middle line:90% to be a way to handle that. 00:20:21.195 --> 00:20:23.520 align:middle line:84% It's as though when you move things, what you do, 00:20:23.520 --> 00:20:25.530 align:middle line:84% really, is take something that you've already 00:20:25.530 --> 00:20:27.720 align:middle line:84% merged in the tree and you merge it again. 00:20:27.720 --> 00:20:29.220 align:middle line:84% And then there have to be principles 00:20:29.220 --> 00:20:33.870 align:middle line:84% that explain where you pronounce it when you merge it 00:20:33.870 --> 00:20:34.650 align:middle line:90% more than once. 00:20:34.650 --> 00:20:37.590 align:middle line:84% Do you pronounce this "what Mary will eat?"-- 00:20:37.590 --> 00:20:39.060 align:middle line:84% "I don't know what Mary will eat." 00:20:39.060 --> 00:20:42.360 align:middle line:84% Do you pronounce it, "I don't know, Mary will eat what"? 00:20:42.360 --> 00:20:45.060 align:middle line:84% We've talked about the fact that some languages have wh-movement 00:20:45.060 --> 00:20:46.050 align:middle line:90% and some don't. 00:20:46.050 --> 00:20:47.970 align:middle line:84% So maybe there are languages that 00:20:47.970 --> 00:20:50.187 align:middle line:84% pronounce the higher copy of "what," that's English, 00:20:50.187 --> 00:20:52.770 align:middle line:84% and languages that pronounce the lower copy of "what," that's, 00:20:52.770 --> 00:20:55.680 align:middle line:90% like, Mandarin, say. 00:20:55.680 --> 00:20:58.302 align:middle line:84% And similarly for reconstruction, 00:20:58.302 --> 00:20:59.760 align:middle line:84% the way to talk about this is yeah, 00:20:59.760 --> 00:21:01.385 align:middle line:84% there are two copies of "what," and you 00:21:01.385 --> 00:21:04.170 align:middle line:84% have to pick one to be the one that you'll interpret. 00:21:04.170 --> 00:21:08.140 align:middle line:84% And the same deal with "two guards," or whatever all else. 00:21:08.140 --> 00:21:10.080 align:middle line:90% So "what" is now in two places. 00:21:10.080 --> 00:21:11.340 align:middle line:90% Yeah? 00:21:11.340 --> 00:21:13.500 align:middle line:84% AUDIENCE: Do you always pronounce the higher? 00:21:13.500 --> 00:21:17.160 align:middle line:84% NORVIN RICHARDS: So one theory about what's going on 00:21:17.160 --> 00:21:20.040 align:middle line:84% is in English, you pronounce the higher 00:21:20.040 --> 00:21:23.010 align:middle line:84% of two copies when you're doing wh- movements. 00:21:23.010 --> 00:21:25.890 align:middle line:84% But maybe the-- so we've said there 00:21:25.890 --> 00:21:28.800 align:middle line:84% are languages in which the way to ask "What did Mary eat?" 00:21:28.800 --> 00:21:31.500 align:middle line:84% is to say literally, "Mary ate what." 00:21:31.500 --> 00:21:33.390 align:middle line:84% So Mandarin is like that, for example, 00:21:33.390 --> 00:21:35.887 align:middle line:90% or Japanese is like that. 00:21:35.887 --> 00:21:37.470 align:middle line:84% There are lots of languages like that. 00:21:37.470 --> 00:21:39.345 align:middle line:84% And maybe a way to talk about those languages 00:21:39.345 --> 00:21:41.070 align:middle line:84% is as far as movement goes, they're 00:21:41.070 --> 00:21:42.930 align:middle line:84% doing all the same movement that English is. 00:21:42.930 --> 00:21:45.408 align:middle line:84% They're just pronouncing the lower copy. 00:21:45.408 --> 00:21:47.200 align:middle line:84% And English is pronouncing the higher copy, 00:21:47.200 --> 00:21:49.140 align:middle line:84% and those are languages that pronounce the lower copy. 00:21:49.140 --> 00:21:51.015 align:middle line:84% That's one way to talk about what's going on. 00:21:51.015 --> 00:21:53.482 align:middle line:90% 00:21:53.482 --> 00:21:54.940 align:middle line:84% So copy theory of movement is meant 00:21:54.940 --> 00:21:59.620 align:middle line:84% to be responsible for these data about reconstruction, which 00:21:59.620 --> 00:22:03.010 align:middle line:84% really seem to make it look like it's misleading to say when you 00:22:03.010 --> 00:22:04.540 align:middle line:90% move something-- 00:22:04.540 --> 00:22:06.730 align:middle line:84% we talk about it as movement because, well, we're 00:22:06.730 --> 00:22:09.230 align:middle line:84% pronouncing it over here and not over there. 00:22:09.230 --> 00:22:11.830 align:middle line:84% And that led us to think of it as now it's here 00:22:11.830 --> 00:22:12.955 align:middle line:90% and it's not there anymore. 00:22:12.955 --> 00:22:14.830 align:middle line:84% Reconstruction makes it look like no, there's 00:22:14.830 --> 00:22:16.480 align:middle line:84% some sense in which it's simultaneously 00:22:16.480 --> 00:22:18.070 align:middle line:84% in all the places it's ever been, 00:22:18.070 --> 00:22:22.132 align:middle line:84% and you get to choose which of the places to interpret it in. 00:22:22.132 --> 00:22:23.590 align:middle line:84% And maybe there's something similar 00:22:23.590 --> 00:22:27.280 align:middle line:84% going on with pronunciation, as I just said. 00:22:27.280 --> 00:22:29.250 align:middle line:84% I don't know if anybody remembers Finnish vowel 00:22:29.250 --> 00:22:30.000 align:middle line:90% harmony. 00:22:30.000 --> 00:22:32.748 align:middle line:90% We're doing nostalgia today. 00:22:32.748 --> 00:22:34.290 align:middle line:84% So I've asked you to remember syntax, 00:22:34.290 --> 00:22:35.873 align:middle line:84% I've asked you to remember morphology. 00:22:35.873 --> 00:22:39.390 align:middle line:84% Now remember this thing that we did very fast about phonology. 00:22:39.390 --> 00:22:41.460 align:middle line:84% So there's this observation about Finnish 00:22:41.460 --> 00:22:46.290 align:middle line:84% that properties like front or back for vowels in Finnish 00:22:46.290 --> 00:22:49.470 align:middle line:84% are just constant across the word, more or less. 00:22:49.470 --> 00:22:52.393 align:middle line:90% So there are words in-- 00:22:52.393 --> 00:22:54.060 align:middle line:84% Finnish has what's called vowel harmony. 00:22:54.060 --> 00:22:56.550 align:middle line:84% There are words in which all the vowels are back 00:22:56.550 --> 00:22:59.202 align:middle line:84% and words in which all the vowels are front. 00:22:59.202 --> 00:23:01.410 align:middle line:84% So there are words like [SPEAKING FINNISH],, which is 00:23:01.410 --> 00:23:02.850 align:middle line:90% the Finnish word for "table." 00:23:02.850 --> 00:23:06.090 align:middle line:84% And if you want to put that in the case that expresses 00:23:06.090 --> 00:23:08.100 align:middle line:84% the meaning "on the table," there's 00:23:08.100 --> 00:23:11.070 align:middle line:84% a suffix, which, when you add it to [SPEAKING FINNISH],, 00:23:11.070 --> 00:23:14.040 align:middle line:84% is pronounced lah, with an ah with the front vowel 00:23:14.040 --> 00:23:16.470 align:middle line:84% because [SPEAKING FINNISH] has front vowels. 00:23:16.470 --> 00:23:18.870 align:middle line:84% And so when you add "la" to it, you get lah. 00:23:18.870 --> 00:23:22.410 align:middle line:90% 00:23:22.410 --> 00:23:25.140 align:middle line:84% If you added la to the word for "chair," 00:23:25.140 --> 00:23:28.110 align:middle line:84% which is [SPEAKING FINNISH],, you'd get [SPEAKING FINNISH].. 00:23:28.110 --> 00:23:30.420 align:middle line:84% So the vowel would be back in the la. 00:23:30.420 --> 00:23:34.410 align:middle line:84% And the idea that I floated, is an old idea in phonology, 00:23:34.410 --> 00:23:37.080 align:middle line:84% is that we should think of properties like front 00:23:37.080 --> 00:23:41.500 align:middle line:84% or back in Finnish as being smeared across the whole word. 00:23:41.500 --> 00:23:43.668 align:middle line:84% So it's not like for every Finnish vowel 00:23:43.668 --> 00:23:45.210 align:middle line:84% in a Finnish word, you get to specify 00:23:45.210 --> 00:23:46.445 align:middle line:90% whether it's front or back. 00:23:46.445 --> 00:23:48.570 align:middle line:84% Front or back is just a property of the whole word. 00:23:48.570 --> 00:23:50.310 align:middle line:84% It's in all of these places at once, 00:23:50.310 --> 00:23:53.130 align:middle line:84% this property of being front or back, 00:23:53.130 --> 00:23:57.120 align:middle line:84% and it appears in more than one place at once. 00:23:57.120 --> 00:23:59.940 align:middle line:84% What I'm talking about now, copy theory of movement, 00:23:59.940 --> 00:24:02.190 align:middle line:90% is like that for syntax. 00:24:02.190 --> 00:24:04.650 align:middle line:84% There are things that are in more than one place 00:24:04.650 --> 00:24:08.670 align:middle line:84% at the same time because you merged them to the tree 00:24:08.670 --> 00:24:10.680 align:middle line:90% over and over again. 00:24:10.680 --> 00:24:13.740 align:middle line:84% And then it is the job of the phonology and the semantics 00:24:13.740 --> 00:24:17.580 align:middle line:84% to take that property of these syntactic representations 00:24:17.580 --> 00:24:20.130 align:middle line:84% and figure out what the heck to do with them. 00:24:20.130 --> 00:24:25.890 align:middle line:84% So syntax is handing phonology structures that literally look 00:24:25.890 --> 00:24:27.442 align:middle line:90% like "What Mary will eat what?" 00:24:27.442 --> 00:24:29.400 align:middle line:84% And phonology has to look at that and go, what? 00:24:29.400 --> 00:24:30.328 align:middle line:90% Wait, which of these? 00:24:30.328 --> 00:24:31.620 align:middle line:90% Where am I going to pronounce-- 00:24:31.620 --> 00:24:33.642 align:middle line:84% I have to pronounce "what" once, and I 00:24:33.642 --> 00:24:35.850 align:middle line:84% have to choose which of these places to pronounce it. 00:24:35.850 --> 00:24:37.808 align:middle line:84% And maybe semantics is doing something similar. 00:24:37.808 --> 00:24:40.590 align:middle line:84% Maybe that's what all this talk about reconstruction is doing. 00:24:40.590 --> 00:24:42.613 align:middle line:84% Semantics similarly looks at that representation 00:24:42.613 --> 00:24:44.280 align:middle line:84% and says, oh, I've got this thing that's 00:24:44.280 --> 00:24:45.238 align:middle line:90% in more than one place. 00:24:45.238 --> 00:24:48.810 align:middle line:84% I can only interpret it in one place, and I'll pick one. 00:24:48.810 --> 00:24:52.230 align:middle line:84% And then big literature figuring out how do we pick? 00:24:52.230 --> 00:24:54.450 align:middle line:84% What are the rules about where we pick? 00:24:54.450 --> 00:24:56.880 align:middle line:90% Lots of work on this. 00:24:56.880 --> 00:24:57.970 align:middle line:90% Copy theory of movement. 00:24:57.970 --> 00:25:02.060 align:middle line:90% 00:25:02.060 --> 00:25:05.090 align:middle line:84% So yeah, reconstruction involves picking a particular position 00:25:05.090 --> 00:25:07.550 align:middle line:84% to interpret a move phrasing, both for binding theory 00:25:07.550 --> 00:25:11.070 align:middle line:90% and for quantifiers. 00:25:11.070 --> 00:25:12.510 align:middle line:90% Topic shift. 00:25:12.510 --> 00:25:16.410 align:middle line:84% I think I can promise that I will never again talk 00:25:16.410 --> 00:25:18.692 align:middle line:90% about guards and buildings. 00:25:18.692 --> 00:25:21.150 align:middle line:84% So if anybody has any questions about guards and buildings, 00:25:21.150 --> 00:25:21.650 align:middle line:90% go ahead. 00:25:21.650 --> 00:25:23.290 align:middle line:90% Get them out of your system. 00:25:23.290 --> 00:25:24.960 align:middle line:84% And we're going to leave them behind. 00:25:24.960 --> 00:25:28.030 align:middle line:84% No more guards, no more buildings. 00:25:28.030 --> 00:25:30.360 align:middle line:90% Let's talk about ellipsis again. 00:25:30.360 --> 00:25:33.000 align:middle line:90% We talked about ellipses before. 00:25:33.000 --> 00:25:39.360 align:middle line:84% So this is a property of sentences where you can leave 00:25:39.360 --> 00:25:41.180 align:middle line:90% out parts of the sentence. 00:25:41.180 --> 00:25:43.680 align:middle line:84% Language is varied with respect to how much of this they do. 00:25:43.680 --> 00:25:47.940 align:middle line:84% English has VP-ellipsis, which is comparatively exotic. 00:25:47.940 --> 00:25:50.070 align:middle line:84% There are other languages that has VP-ellipsis, 00:25:50.070 --> 00:25:52.210 align:middle line:84% but it's certainly not every language. 00:25:52.210 --> 00:25:54.510 align:middle line:84% So VP-ellipsis is this option that English 00:25:54.510 --> 00:25:56.730 align:middle line:84% and some other languages have where 00:25:56.730 --> 00:25:58.608 align:middle line:84% if you're about to repeat a verb phrase, 00:25:58.608 --> 00:26:00.900 align:middle line:84% you have the option of just leaving out the second verb 00:26:00.900 --> 00:26:01.400 align:middle line:90% phrase. 00:26:01.400 --> 00:26:03.930 align:middle line:84% So you can say things like, "Adam ate an apple and Eve did, 00:26:03.930 --> 00:26:07.245 align:middle line:84% too," where the missing verb phrase is "ate an apple." 00:26:07.245 --> 00:26:11.050 align:middle line:90% 00:26:11.050 --> 00:26:13.153 align:middle line:84% There's another phenomenon which is a lot more 00:26:13.153 --> 00:26:14.320 align:middle line:90% cross-linguistically common. 00:26:14.320 --> 00:26:19.270 align:middle line:84% It's called sluicing, where you get to leave out 00:26:19.270 --> 00:26:22.280 align:middle line:90% most of a question. 00:26:22.280 --> 00:26:25.220 align:middle line:84% So you can say things like, "Adam ate something, 00:26:25.220 --> 00:26:27.670 align:middle line:84% but I don't know what" where this means 00:26:27.670 --> 00:26:30.160 align:middle line:90% "I don't know what Adam ate." 00:26:30.160 --> 00:26:33.002 align:middle line:84% So what's been left out is the rest of the question, 00:26:33.002 --> 00:26:34.210 align:middle line:90% "I don't know what Adam ate." 00:26:34.210 --> 00:26:37.540 align:middle line:84% All you pronounce is the "what" part. 00:26:37.540 --> 00:26:39.550 align:middle line:90% This is called sluicing. 00:26:39.550 --> 00:26:40.900 align:middle line:90% There's an old name for it. 00:26:40.900 --> 00:26:43.973 align:middle line:90% It's called that because-- 00:26:43.973 --> 00:26:45.640 align:middle line:84% I think I've made this mistake sometimes 00:26:45.640 --> 00:26:50.710 align:middle line:84% on slides where something that's called-- 00:26:50.710 --> 00:26:53.470 align:middle line:84% we've been calling TP, which people these days usually 00:26:53.470 --> 00:26:54.820 align:middle line:90% do call TP. 00:26:54.820 --> 00:26:57.305 align:middle line:90% It used to be called IP. 00:26:57.305 --> 00:26:58.430 align:middle line:90% It's just a different name. 00:26:58.430 --> 00:26:59.440 align:middle line:84% So now we call it a tense phrase. 00:26:59.440 --> 00:27:01.240 align:middle line:84% It used to be called an inflection phrase. 00:27:01.240 --> 00:27:03.910 align:middle line:84% And we've had some slides where I accidentally left in an IP 00:27:03.910 --> 00:27:06.490 align:middle line:84% and I apologized, and had to go in and fix it 00:27:06.490 --> 00:27:07.667 align:middle line:90% before I posted the slides. 00:27:07.667 --> 00:27:10.000 align:middle line:84% I hope I managed to do that for all the slides I posted. 00:27:10.000 --> 00:27:16.690 align:middle line:84% Before it was called an IP, it was called an S. So if you-- 00:27:16.690 --> 00:27:18.355 align:middle line:90% S stood for sentence. 00:27:18.355 --> 00:27:21.040 align:middle line:90% 00:27:21.040 --> 00:27:24.340 align:middle line:84% So it was before our terminology got as sophisticated 00:27:24.340 --> 00:27:25.400 align:middle line:90% as it is today. 00:27:25.400 --> 00:27:27.250 align:middle line:90% So S stood for sentence. 00:27:27.250 --> 00:27:30.370 align:middle line:84% And sluicing was a cute name for this. 00:27:30.370 --> 00:27:36.940 align:middle line:84% The idea was when you say "I don't know what Adam ate," 00:27:36.940 --> 00:27:38.920 align:middle line:84% you're getting rid of "Adam ate." 00:27:38.920 --> 00:27:44.050 align:middle line:84% This is the S. So you are "S-loosing." 00:27:44.050 --> 00:27:45.520 align:middle line:84% You're getting rid of the S. Yeah, 00:27:45.520 --> 00:27:47.410 align:middle line:84% so that's how it got the name "sluicing." 00:27:47.410 --> 00:27:48.730 align:middle line:90% Yeah, it's a cute name. 00:27:48.730 --> 00:27:51.160 align:middle line:84% I'm sorry, that's the only excuse I have for telling you 00:27:51.160 --> 00:27:53.380 align:middle line:84% that this thing used to be called S. That, 00:27:53.380 --> 00:27:56.830 align:middle line:84% and that if you ever read ancient literature in syntax, 00:27:56.830 --> 00:27:59.590 align:middle line:84% you'll sometimes see it called S. 00:27:59.590 --> 00:28:01.150 align:middle line:90% So VP-ellipsis. 00:28:01.150 --> 00:28:02.230 align:middle line:90% Fairly common phenomenon. 00:28:02.230 --> 00:28:04.510 align:middle line:84% Sluicing, possibly a universal phenomenon. 00:28:04.510 --> 00:28:06.940 align:middle line:84% There's a lot of work on whether things 00:28:06.940 --> 00:28:08.440 align:middle line:84% that look like sluicing really are 00:28:08.440 --> 00:28:12.280 align:middle line:90% sluicing in various languages. 00:28:12.280 --> 00:28:15.190 align:middle line:84% And at least one way of understanding what's going on 00:28:15.190 --> 00:28:16.453 align:middle line:90% here is that these are-- 00:28:16.453 --> 00:28:18.370 align:middle line:84% so there are a couple of ways of understanding 00:28:18.370 --> 00:28:20.453 align:middle line:84% what's going on here, and we will talk about them. 00:28:20.453 --> 00:28:21.910 align:middle line:84% But here's one way you could say, 00:28:21.910 --> 00:28:26.080 align:middle line:84% yeah, it's possible sometimes to just refrain from pronouncing 00:28:26.080 --> 00:28:27.560 align:middle line:90% parts of a sentence. 00:28:27.560 --> 00:28:33.640 align:middle line:84% So if you have built a sentence with the structure "Adam 00:28:33.640 --> 00:28:36.340 align:middle line:84% ate something, but I don't know what Adam ate," well, 00:28:36.340 --> 00:28:37.610 align:middle line:90% you can say that. 00:28:37.610 --> 00:28:39.820 align:middle line:84% But you also have the option of refraining 00:28:39.820 --> 00:28:43.342 align:middle line:84% from repeating the "Adam ate" at the end there. 00:28:43.342 --> 00:28:44.425 align:middle line:90% You can just leave it out. 00:28:44.425 --> 00:28:48.000 align:middle line:90% 00:28:48.000 --> 00:28:49.440 align:middle line:90% That's one way to talk about it. 00:28:49.440 --> 00:28:51.690 align:middle line:84% There's another way to talk about it, of course, which 00:28:51.690 --> 00:28:54.930 align:middle line:84% is to say, no, look, if you say "Adam ate something, 00:28:54.930 --> 00:28:58.110 align:middle line:84% but I don't know what," well, you have said, 00:28:58.110 --> 00:29:00.990 align:middle line:84% "Adam ate something but I don't know what." 00:29:00.990 --> 00:29:03.270 align:middle line:84% We need a structure for that where 00:29:03.270 --> 00:29:05.130 align:middle line:90% there's no sentence after what. 00:29:05.130 --> 00:29:06.570 align:middle line:90% There's just nothing there. 00:29:06.570 --> 00:29:09.690 align:middle line:84% And yeah, it's the job of the semantics 00:29:09.690 --> 00:29:12.690 align:middle line:84% to interpret that as "I don't know what Adam ate." 00:29:12.690 --> 00:29:16.480 align:middle line:84% But the syntax is not building a structure, 00:29:16.480 --> 00:29:19.053 align:middle line:84% "I don't know what Adam ate," and then refraining 00:29:19.053 --> 00:29:20.220 align:middle line:90% from pronouncing part of it. 00:29:20.220 --> 00:29:22.380 align:middle line:84% There's none of this ellipsis stuff. 00:29:22.380 --> 00:29:25.170 align:middle line:84% You're just-- we have a pretty sophisticated way 00:29:25.170 --> 00:29:28.823 align:middle line:84% of interpreting things like "I don't know what." 00:29:28.823 --> 00:29:30.990 align:middle line:84% Do people see the difference between these theories? 00:29:30.990 --> 00:29:32.940 align:middle line:84% And one of them says the syntax always 00:29:32.940 --> 00:29:35.430 align:middle line:84% creates complete sentences, and then the phonology 00:29:35.430 --> 00:29:38.430 align:middle line:84% has the option of refraining from pronouncing parts of them. 00:29:38.430 --> 00:29:39.330 align:middle line:90% Yeah. 00:29:39.330 --> 00:29:41.250 align:middle line:84% The other says, no, the syntax has the option 00:29:41.250 --> 00:29:43.380 align:middle line:84% of creating things that are not complete sentences, 00:29:43.380 --> 00:29:47.070 align:middle line:84% and then the semantics just has to deal, figure out 00:29:47.070 --> 00:29:49.950 align:middle line:90% how to interpret those things. 00:29:49.950 --> 00:29:51.540 align:middle line:84% Those are two takes that people have 00:29:51.540 --> 00:29:54.210 align:middle line:90% had on this kind of phenomena. 00:29:54.210 --> 00:29:55.440 align:middle line:90% Yeah? 00:29:55.440 --> 00:29:57.690 align:middle line:84% Now let me give you one reason to take one of these 00:29:57.690 --> 00:30:00.063 align:middle line:84% takes seriously, more seriously than the other one. 00:30:00.063 --> 00:30:01.980 align:middle line:84% Though, they are both live options, and people 00:30:01.980 --> 00:30:03.810 align:middle line:90% argue for both of them. 00:30:03.810 --> 00:30:04.600 align:middle line:90% Different people. 00:30:04.600 --> 00:30:06.975 align:middle line:84% Some people argue for one, other people argue for others. 00:30:06.975 --> 00:30:09.330 align:middle line:84% Most people do not argue for both. 00:30:09.330 --> 00:30:12.010 align:middle line:90% Here's one. 00:30:12.010 --> 00:30:14.320 align:middle line:84% We've talked about preposition stranding. 00:30:14.320 --> 00:30:17.650 align:middle line:84% So preposition stranding is this fact 00:30:17.650 --> 00:30:21.010 align:middle line:84% that in certain strange and exotic languages like the one 00:30:21.010 --> 00:30:24.400 align:middle line:84% I'm speaking, you have the option 00:30:24.400 --> 00:30:27.040 align:middle line:84% of asking two kinds of questions if you 00:30:27.040 --> 00:30:30.280 align:middle line:84% want to ask a question about the object of a preposition. 00:30:30.280 --> 00:30:33.550 align:middle line:84% You can ask questions like "Who was he talking with?" 00:30:33.550 --> 00:30:37.750 align:middle line:84% Or you can ask questions like "With whom was he talking?" 00:30:37.750 --> 00:30:40.450 align:middle line:84% That is, if you want to ask a question about the object 00:30:40.450 --> 00:30:43.960 align:middle line:84% of "with," you can just wh-move the object of "with," 00:30:43.960 --> 00:30:46.810 align:middle line:84% and leave the "with" behind: "Who was he talking with?" 00:30:46.810 --> 00:30:49.510 align:middle line:84% Or you can move the entire prepositional phrase: 00:30:49.510 --> 00:30:51.250 align:middle line:90% "With whom was he talking?" 00:30:51.250 --> 00:30:55.430 align:middle line:84% I, at least, would much rather say the first of those. 00:30:55.430 --> 00:30:58.600 align:middle line:84% But some people are able to say the second one. 00:30:58.600 --> 00:31:00.340 align:middle line:84% It depends on how savagely you were 00:31:00.340 --> 00:31:02.620 align:middle line:84% beaten by English teachers in high school. 00:31:02.620 --> 00:31:05.665 align:middle line:84% So your English teachers in high school 00:31:05.665 --> 00:31:08.290 align:middle line:84% maybe tried to convince you that the right way to speak English 00:31:08.290 --> 00:31:10.480 align:middle line:84% involves the sentence on the right. 00:31:10.480 --> 00:31:13.030 align:middle line:84% They did that-- I mean, they had the best of intentions. 00:31:13.030 --> 00:31:15.730 align:middle line:84% They did that because most languages only 00:31:15.730 --> 00:31:17.510 align:middle line:90% have that second option. 00:31:17.510 --> 00:31:20.260 align:middle line:84% So English is quite unusual in having that first option, 00:31:20.260 --> 00:31:25.540 align:middle line:84% and scholars of the English language sometime in the 1700s 00:31:25.540 --> 00:31:27.820 align:middle line:84% decided that English would be much cooler if it 00:31:27.820 --> 00:31:29.320 align:middle line:90% were like every other language. 00:31:29.320 --> 00:31:31.790 align:middle line:84% It's basically grammatical peer pressure. 00:31:31.790 --> 00:31:33.490 align:middle line:84% So they were like, wouldn't it be nice 00:31:33.490 --> 00:31:36.340 align:middle line:84% if English were more like Latin, and French, 00:31:36.340 --> 00:31:38.320 align:middle line:84% and other civilized languages where you can't 00:31:38.320 --> 00:31:40.752 align:middle line:90% leave prepositions behind? 00:31:40.752 --> 00:31:43.210 align:middle line:84% So there are some languages that have this option of what's 00:31:43.210 --> 00:31:45.190 align:middle line:84% called preposition stranding, leaving "with" 00:31:45.190 --> 00:31:50.680 align:middle line:84% behind and wh-moving the object of the preposition out, as 00:31:50.680 --> 00:31:52.060 align:middle line:90% in "Who is he talking with?" 00:31:52.060 --> 00:31:53.870 align:middle line:84% Most languages don't have that option. 00:31:53.870 --> 00:31:55.480 align:middle line:84% You only have the option on the right. 00:31:55.480 --> 00:31:57.830 align:middle line:84% This is all-- we've said all of this. 00:31:57.830 --> 00:31:59.650 align:middle line:84% So I say we should be proud of the fact 00:31:59.650 --> 00:32:02.050 align:middle line:84% that we can say, "Who was he talking with?" 00:32:02.050 --> 00:32:03.520 align:middle line:90% It should be on our currency. 00:32:03.520 --> 00:32:06.130 align:middle line:84% It should be part of the official motto of the US. 00:32:06.130 --> 00:32:07.210 align:middle line:90% "Who are you talking to?" 00:32:07.210 --> 00:32:08.920 align:middle line:90% That should be our motto. 00:32:08.920 --> 00:32:11.260 align:middle line:90% No? 00:32:11.260 --> 00:32:15.140 align:middle line:84% Now as I said, most languages are not like this. 00:32:15.140 --> 00:32:17.110 align:middle line:84% So here are some Russian examples. 00:32:17.110 --> 00:32:19.540 align:middle line:84% In Russian, this is Russian for "Peter 00:32:19.540 --> 00:32:24.040 align:middle line:84% was talking with somebody but I don't know whom." 00:32:24.040 --> 00:32:26.200 align:middle line:84% And similarly, let's concentrate first 00:32:26.200 --> 00:32:27.430 align:middle line:90% on the sentences at the end. 00:32:27.430 --> 00:32:30.340 align:middle line:84% In Russian, Russian is like most civilized languages in which 00:32:30.340 --> 00:32:32.350 align:middle line:84% you cannot say, "Who is he talking with?" 00:32:32.350 --> 00:32:35.080 align:middle line:84% You can't say, [SPEAKING RUSSIAN].. 00:32:35.080 --> 00:32:37.120 align:middle line:84% You have to say [SPEAKING RUSSIAN].. 00:32:37.120 --> 00:32:39.130 align:middle line:90% So "With whom was he talking?" 00:32:39.130 --> 00:32:40.750 align:middle line:84% So that the preposition "with," which 00:32:40.750 --> 00:32:42.370 align:middle line:84% is [SPEAKING RUSSIAN] in Russian. 00:32:42.370 --> 00:32:44.290 align:middle line:84% Russian has longer prepositions than that, 00:32:44.290 --> 00:32:46.420 align:middle line:84% but that's one of them, prepositions 00:32:46.420 --> 00:32:48.605 align:middle line:90% that are just a consonant. 00:32:48.605 --> 00:32:50.230 align:middle line:84% You have to take that preposition along 00:32:50.230 --> 00:32:53.260 align:middle line:90% with the wh- word. 00:32:53.260 --> 00:32:58.650 align:middle line:84% And relatedly, possibly, Russian has sluicing. 00:32:58.650 --> 00:33:01.170 align:middle line:84% But in Russian-- so let's go back to English. 00:33:01.170 --> 00:33:03.760 align:middle line:84% In English, if you do sluicing, you have two options. 00:33:03.760 --> 00:33:05.968 align:middle line:84% You can either say, "Peter was talking with somebody, 00:33:05.968 --> 00:33:09.760 align:middle line:84% but I don't know who," leaving the preposition out. 00:33:09.760 --> 00:33:11.760 align:middle line:84% Or you can say, "Peter was talking with someone, 00:33:11.760 --> 00:33:13.270 align:middle line:90% but I don't know with whom." 00:33:13.270 --> 00:33:16.430 align:middle line:90% Yeah, you can say that, too. 00:33:16.430 --> 00:33:17.930 align:middle line:84% Russian only has the second option. 00:33:17.930 --> 00:33:19.190 align:middle line:84% You can only say the Russian version 00:33:19.190 --> 00:33:20.648 align:middle line:84% of "Peter was talking with someone, 00:33:20.648 --> 00:33:23.310 align:middle line:90% but I don't with whom." 00:33:23.310 --> 00:33:26.188 align:middle line:84% You can't say in Russian, "Peter was talking with somebody, 00:33:26.188 --> 00:33:27.105 align:middle line:90% but I don't know who." 00:33:27.105 --> 00:33:29.690 align:middle line:90% 00:33:29.690 --> 00:33:32.810 align:middle line:84% And this is English and Russian, but this is quite widespread. 00:33:32.810 --> 00:33:34.460 align:middle line:90% It's generally true. 00:33:34.460 --> 00:33:36.260 align:middle line:84% There are some interesting counterexamples, 00:33:36.260 --> 00:33:37.930 align:middle line:84% which people get interested in and try 00:33:37.930 --> 00:33:39.180 align:middle line:90% to figure out what's going on. 00:33:39.180 --> 00:33:42.200 align:middle line:84% But it's generally true that if you are like English 00:33:42.200 --> 00:33:44.450 align:middle line:84% in allowing prepositions to be left behind, 00:33:44.450 --> 00:33:47.360 align:middle line:84% you can leave prepositions out in sluicing. 00:33:47.360 --> 00:33:50.270 align:middle line:84% And if you are like Russian in which your prepositions have 00:33:50.270 --> 00:33:52.220 align:middle line:84% to move along with your wh- phrase, 00:33:52.220 --> 00:33:56.030 align:middle line:84% then your prepositions have to be kept under sluicing. 00:33:56.030 --> 00:33:57.590 align:middle line:84% So let's think about what that means 00:33:57.590 --> 00:33:58.882 align:middle line:90% for these theories of sluicing. 00:33:58.882 --> 00:34:01.750 align:middle line:90% 00:34:01.750 --> 00:34:03.830 align:middle line:84% The two theories of sluicing again, one of them-- 00:34:03.830 --> 00:34:05.230 align:middle line:90% let's take the first one first. 00:34:05.230 --> 00:34:07.570 align:middle line:84% One of them said, when you say "Peter was talking with 00:34:07.570 --> 00:34:09.790 align:middle line:84% somebody, but I don't know who," well, 00:34:09.790 --> 00:34:11.074 align:middle line:90% what you're doing is you're-- 00:34:11.074 --> 00:34:13.449 align:middle line:84% the syntax is generating "Peter was talking with someone, 00:34:13.449 --> 00:34:15.520 align:middle line:84% but I don't know who he was talking with." 00:34:15.520 --> 00:34:19.510 align:middle line:84% And then you leave out everything 00:34:19.510 --> 00:34:23.870 align:middle line:84% in the lowest sentence except for the wh- word itself. 00:34:23.870 --> 00:34:26.280 align:middle line:84% So you start off with "Peter was talking with somebody, 00:34:26.280 --> 00:34:28.030 align:middle line:84% but I don't know who he was talking with," 00:34:28.030 --> 00:34:29.590 align:middle line:84% and then you leave out "he was talking with." 00:34:29.590 --> 00:34:31.350 align:middle line:84% So you end up with "I don't know who." 00:34:31.350 --> 00:34:33.100 align:middle line:84% Now that's a story where sluicing involves 00:34:33.100 --> 00:34:35.560 align:middle line:84% creating a complete sentence and then electing 00:34:35.560 --> 00:34:38.380 align:middle line:90% to not pronounce part of it. 00:34:38.380 --> 00:34:41.170 align:middle line:84% If we have that kind of theory of sluicing, 00:34:41.170 --> 00:34:45.730 align:middle line:84% then the facts in English and Russian 00:34:45.730 --> 00:34:49.670 align:middle line:84% boil down to a single difference between the languages. 00:34:49.670 --> 00:34:52.480 align:middle line:84% So the single difference between the languages 00:34:52.480 --> 00:34:55.179 align:middle line:84% has to do with whether you have to take prepositions along 00:34:55.179 --> 00:34:59.458 align:middle line:84% with you when you wh-move the object of a preposition, 00:34:59.458 --> 00:35:01.000 align:middle line:84% whether you have the option of saying 00:35:01.000 --> 00:35:02.330 align:middle line:90% "Who is he talking with?" 00:35:02.330 --> 00:35:04.960 align:middle line:84% So there's a relation between the fact that in English 00:35:04.960 --> 00:35:07.450 align:middle line:84% but not in Russian, you can say, "Who was he talking with?" 00:35:07.450 --> 00:35:09.400 align:middle line:84% And the fact that in English, but not in Russian, 00:35:09.400 --> 00:35:11.233 align:middle line:84% you can say, "Peter was talking to somebody, 00:35:11.233 --> 00:35:13.990 align:middle line:90% but I don't know who." 00:35:13.990 --> 00:35:15.410 align:middle line:90% Those are both the same fact. 00:35:15.410 --> 00:35:18.370 align:middle line:84% The fact is when the syntax is constructing 00:35:18.370 --> 00:35:21.245 align:middle line:84% the sluiced example, in English, you have the option of saying, 00:35:21.245 --> 00:35:22.870 align:middle line:84% "I don't know who he was talking with," 00:35:22.870 --> 00:35:24.880 align:middle line:84% and then leaving out "he was talking with." 00:35:24.880 --> 00:35:26.740 align:middle line:84% In Russian, you don't have that option. 00:35:26.740 --> 00:35:28.870 align:middle line:84% You have to say, "I don't know with whom he was talking." 00:35:28.870 --> 00:35:30.820 align:middle line:84% And that's why when you do sluicing, you have to say, 00:35:30.820 --> 00:35:32.820 align:middle line:84% "but I don't know with whom," and you can't say, 00:35:32.820 --> 00:35:33.820 align:middle line:90% "I don't know who." 00:35:33.820 --> 00:35:35.770 align:middle line:84% So these facts are related to each other. 00:35:35.770 --> 00:35:37.930 align:middle line:84% If you can strand prepositions, then you 00:35:37.930 --> 00:35:39.340 align:middle line:90% can leave them out in sluicing. 00:35:39.340 --> 00:35:42.910 align:middle line:84% And those facts are connected if sluicing involves, well, 00:35:42.910 --> 00:35:45.550 align:middle line:84% actual wh-movement, together with a decision 00:35:45.550 --> 00:35:47.225 align:middle line:84% to fail to pronounce part of a sentence. 00:35:47.225 --> 00:35:49.840 align:middle line:90% 00:35:49.840 --> 00:35:51.803 align:middle line:84% Yep, that's one argument for that approach. 00:35:51.803 --> 00:35:54.220 align:middle line:84% On the other hand, if you have the kind of approach that's 00:35:54.220 --> 00:35:59.080 align:middle line:84% like, no, look, there's no-- you don't always 00:35:59.080 --> 00:36:00.280 align:middle line:90% create complete sentences. 00:36:00.280 --> 00:36:02.697 align:middle line:84% The syntax when you say, "Peter was talking with somebody, 00:36:02.697 --> 00:36:05.860 align:middle line:84% but I don't know who," the syntax is just creating 00:36:05.860 --> 00:36:07.090 align:middle line:90% "I don't know who." 00:36:07.090 --> 00:36:09.130 align:middle line:84% And then it is the job of the semantics 00:36:09.130 --> 00:36:11.500 align:middle line:84% to figure out what the heck you mean when you say that, 00:36:11.500 --> 00:36:14.380 align:middle line:84% and to supply the rest of the sentence from context. 00:36:14.380 --> 00:36:18.910 align:middle line:84% Well then, it's less clear why these facts 00:36:18.910 --> 00:36:21.070 align:middle line:84% should track each other, the fact about whether you 00:36:21.070 --> 00:36:22.778 align:middle line:84% constrain prepositions and the fact about 00:36:22.778 --> 00:36:24.487 align:middle line:84% whether you can leave it out in a sluice, 00:36:24.487 --> 00:36:27.100 align:middle line:84% because the sluice doesn't, on that kind of theory, literally 00:36:27.100 --> 00:36:30.740 align:middle line:90% involve any wh-movement. 00:36:30.740 --> 00:36:33.900 align:middle line:90% Another similar kind of fact. 00:36:33.900 --> 00:36:40.650 align:middle line:84% German has verbs that assign entertaining cases 00:36:40.650 --> 00:36:42.650 align:middle line:90% to their objects. 00:36:42.650 --> 00:36:44.240 align:middle line:84% So English does not have a whole lot 00:36:44.240 --> 00:36:47.690 align:middle line:84% of case morphology, but way back when, we talked a little bit 00:36:47.690 --> 00:36:50.360 align:middle line:84% back in syntax about case morphology, this idea 00:36:50.360 --> 00:36:54.740 align:middle line:84% that nouns have morphology on them telling you roughly what 00:36:54.740 --> 00:36:56.700 align:middle line:90% they're doing in the sentence. 00:36:56.700 --> 00:36:58.283 align:middle line:84% So we talked about the fact that there 00:36:58.283 --> 00:37:00.140 align:middle line:84% are languages that have a marker that 00:37:00.140 --> 00:37:02.837 align:middle line:84% says, hey, I'm the subject, or hey, I'm the object. 00:37:02.837 --> 00:37:04.670 align:middle line:84% There are languages that have other markers, 00:37:04.670 --> 00:37:07.400 align:middle line:84% these case markers, we called them. 00:37:07.400 --> 00:37:11.180 align:middle line:84% German has more of this stuff than English does. 00:37:11.180 --> 00:37:14.240 align:middle line:84% And in particular, German has this phenomenon sometimes 00:37:14.240 --> 00:37:16.728 align:middle line:84% called quirky case, where there are verbs. 00:37:16.728 --> 00:37:18.770 align:middle line:84% One of the fun things about-- one of the many fun 00:37:18.770 --> 00:37:21.590 align:middle line:84% things about learning German is that if you 00:37:21.590 --> 00:37:24.140 align:middle line:84% want to learn a verb like "schmeicheln," which 00:37:24.140 --> 00:37:26.760 align:middle line:84% is the German verb to flatter, first of all, 00:37:26.760 --> 00:37:29.340 align:middle line:84% you must learn to pronounce it possibly better than I just 00:37:29.340 --> 00:37:29.840 align:middle line:90% did. 00:37:29.840 --> 00:37:33.260 align:middle line:84% And second, you must learn that its object is dative. 00:37:33.260 --> 00:37:33.920 align:middle line:90% Why? 00:37:33.920 --> 00:37:36.080 align:middle line:90% Well, it just is. 00:37:36.080 --> 00:37:38.690 align:middle line:84% So part of learning German is learning 00:37:38.690 --> 00:37:42.290 align:middle line:84% that if you flatter someone, you flatter them datively. 00:37:42.290 --> 00:37:45.350 align:middle line:84% To say that the object of "schmeicheln" is dative 00:37:45.350 --> 00:37:46.880 align:middle line:84% is to say that when you want to say 00:37:46.880 --> 00:37:49.430 align:middle line:84% he wants to flatter someone, the word for someone 00:37:49.430 --> 00:37:53.090 align:middle line:84% is "jemandem" and ends in an M, which is the mark 00:37:53.090 --> 00:37:55.700 align:middle line:90% that it's dative. 00:37:55.700 --> 00:37:58.760 align:middle line:84% If you wanted to say, for example, he wants to praise 00:37:58.760 --> 00:38:01.430 align:middle line:84% someone, well, praise, "loben," is 00:38:01.430 --> 00:38:04.550 align:middle line:84% like most transitive verbs in that its object is accusative. 00:38:04.550 --> 00:38:09.020 align:middle line:84% So now someone isn't "jemandem," it's "jemanden." 00:38:09.020 --> 00:38:10.580 align:middle line:90% A fun, quirky fact about German. 00:38:10.580 --> 00:38:12.260 align:middle line:84% If any of you were considering learning German, 00:38:12.260 --> 00:38:14.468 align:middle line:84% this is one of the many things you will get to learn. 00:38:14.468 --> 00:38:15.380 align:middle line:90% Yeah? 00:38:15.380 --> 00:38:16.120 align:middle line:90% Yeah. 00:38:16.120 --> 00:38:20.360 align:middle line:84% Now cool fact, if you sluice in German-- 00:38:20.360 --> 00:38:23.360 align:middle line:84% German's like English, it has sluicing. 00:38:23.360 --> 00:38:25.700 align:middle line:84% If you sluice in German, if you want to say, 00:38:25.700 --> 00:38:28.700 align:middle line:84% "He wants to flatter somebody, but they don't know who," 00:38:28.700 --> 00:38:30.918 align:middle line:90% you can say that in German. 00:38:30.918 --> 00:38:32.960 align:middle line:84% If you want to say, "He wants to praise somebody, 00:38:32.960 --> 00:38:36.780 align:middle line:84% but they don't know who," you can say that in German, too. 00:38:36.780 --> 00:38:39.630 align:middle line:84% But if you say, "He wants to flatter somebody, 00:38:39.630 --> 00:38:42.065 align:middle line:84% but they don't know who," well, the word for "who," 00:38:42.065 --> 00:38:43.440 align:middle line:84% it's like the word for "someone." 00:38:43.440 --> 00:38:46.020 align:middle line:84% It has different dative and accusative forms. 00:38:46.020 --> 00:38:48.930 align:middle line:84% So as you can see in the slide, "He wants to flatter someone, 00:38:48.930 --> 00:38:51.900 align:middle line:84% but they don't know who," the word for "who" has to be "wem". 00:38:51.900 --> 00:38:53.970 align:middle line:90% Has to be dative. 00:38:53.970 --> 00:38:55.680 align:middle line:84% Whereas "He wants to praise somebody, 00:38:55.680 --> 00:38:58.440 align:middle line:84% but they don't know who," the word for "who" has to be "wen," 00:38:58.440 --> 00:39:01.003 align:middle line:90% has to be accusative. 00:39:01.003 --> 00:39:03.420 align:middle line:84% So let's think about what that means on the two approaches 00:39:03.420 --> 00:39:04.415 align:middle line:90% to sluicing. 00:39:04.415 --> 00:39:06.540 align:middle line:84% On the approach to sluicing where sluicing involves 00:39:06.540 --> 00:39:08.760 align:middle line:84% creating a complete sentence and then 00:39:08.760 --> 00:39:11.370 align:middle line:84% forgetting to say part of it, just one 00:39:11.370 --> 00:39:14.730 align:middle line:84% theory of sluicing we talked about, well, this makes sense. 00:39:14.730 --> 00:39:17.940 align:middle line:84% Because that first sluicing example, 00:39:17.940 --> 00:39:20.970 align:middle line:84% you're saying, in German, "He wants to flatter somebody, 00:39:20.970 --> 00:39:23.640 align:middle line:84% but they don't know who he wants to flatter." 00:39:23.640 --> 00:39:26.153 align:middle line:84% And "who" starts off as the object of "schmeicheln," 00:39:26.153 --> 00:39:28.320 align:middle line:84% a version of "schmeicheln," that you're not actually 00:39:28.320 --> 00:39:29.880 align:middle line:90% going to say. 00:39:29.880 --> 00:39:33.330 align:middle line:84% And because it's the object of "schmeicheln," it is dative, 00:39:33.330 --> 00:39:36.513 align:middle line:84% and so it gets pronounced with dative. 00:39:36.513 --> 00:39:38.930 align:middle line:84% And then if you want to say, "He wants to praise somebody, 00:39:38.930 --> 00:39:41.513 align:middle line:84% but they don't know who," well, you start off by constructing, 00:39:41.513 --> 00:39:43.347 align:middle line:84% "He wants to praise somebody, but they don't 00:39:43.347 --> 00:39:44.968 align:middle line:90% know who he wants to praise." 00:39:44.968 --> 00:39:46.760 align:middle line:84% "Who" starts off as the object of "praise," 00:39:46.760 --> 00:39:49.160 align:middle line:90% and so it's accusative. 00:39:49.160 --> 00:39:51.440 align:middle line:84% And so we have an account of why these words for "who" 00:39:51.440 --> 00:39:54.110 align:middle line:84% are dative or accusative, depending on the properties 00:39:54.110 --> 00:39:57.080 align:middle line:90% of a verb which you cannot hear. 00:39:57.080 --> 00:40:01.850 align:middle line:84% The idea is yeah, you can't hear it, but it's there. 00:40:01.850 --> 00:40:03.440 align:middle line:90% And it does its thing. 00:40:03.440 --> 00:40:05.468 align:middle line:84% It makes "who" either dative or accusative, 00:40:05.468 --> 00:40:07.010 align:middle line:84% and then you move "who" out of there, 00:40:07.010 --> 00:40:09.510 align:middle line:84% and then you forget to pronounce the verb. 00:40:09.510 --> 00:40:12.147 align:middle line:90% Yeah, that's how sluicing works. 00:40:12.147 --> 00:40:14.230 align:middle line:84% On the alternative approach to sluicing that says, 00:40:14.230 --> 00:40:16.480 align:middle line:84% no, it's just semanticists are smarter 00:40:16.480 --> 00:40:19.348 align:middle line:84% than you think they are, you can give them, 00:40:19.348 --> 00:40:21.640 align:middle line:84% "He wants to flatter someone, but they don't know who," 00:40:21.640 --> 00:40:23.260 align:middle line:90% and it will be their job-- 00:40:23.260 --> 00:40:25.703 align:middle line:84% that will just have all the structure that you can hear. 00:40:25.703 --> 00:40:27.370 align:middle line:84% There's no complete sentence down there. 00:40:27.370 --> 00:40:31.750 align:middle line:84% The end of the sentence is just "who." 00:40:31.750 --> 00:40:34.480 align:middle line:84% Well, the semanticists will have to be so smart that they 00:40:34.480 --> 00:40:38.410 align:middle line:84% can know that the verb "to flatter" in German 00:40:38.410 --> 00:40:40.030 align:middle line:84% assigns dative case to its object, 00:40:40.030 --> 00:40:41.530 align:middle line:84% even though "who" is, in some sense, 00:40:41.530 --> 00:40:43.630 align:middle line:84% not the object of "flatter" on that story. 00:40:43.630 --> 00:40:46.540 align:middle line:84% It's never been anywhere near the verb "flatter." 00:40:46.540 --> 00:40:49.570 align:middle line:84% So there are some complications for that approach 00:40:49.570 --> 00:40:50.710 align:middle line:90% in these kinds of facts. 00:40:50.710 --> 00:40:53.240 align:middle line:90% 00:40:53.240 --> 00:40:57.380 align:middle line:84% So those are two reasons to take seriously the idea 00:40:57.380 --> 00:41:02.650 align:middle line:84% that ellipsis, this has been concentrating on sluicing, 00:41:02.650 --> 00:41:07.870 align:middle line:84% is a process in which you create a complete sentence, 00:41:07.870 --> 00:41:09.670 align:middle line:84% a syntactically present structure. 00:41:09.670 --> 00:41:13.360 align:middle line:84% When you say "John is eating something, 00:41:13.360 --> 00:41:15.235 align:middle line:84% but I don't know what," you really are saying 00:41:15.235 --> 00:41:16.818 align:middle line:84% "John is eating something, but I don't 00:41:16.818 --> 00:41:18.040 align:middle line:90% know what John is eating." 00:41:18.040 --> 00:41:21.520 align:middle line:84% And then you are leaving out "John is eating" 00:41:21.520 --> 00:41:23.410 align:middle line:90% from the end of the sentence. 00:41:23.410 --> 00:41:25.090 align:middle line:84% So there's this process of ellipsis 00:41:25.090 --> 00:41:28.660 align:middle line:84% that gets rid of stuff that's syntactically present so 00:41:28.660 --> 00:41:30.070 align:middle line:90% that it's not pronounced. 00:41:30.070 --> 00:41:32.110 align:middle line:90% But it is interpreted. 00:41:32.110 --> 00:41:33.490 align:middle line:84% And it can have effects on things 00:41:33.490 --> 00:41:35.380 align:middle line:90% like case, as we saw in German. 00:41:35.380 --> 00:41:37.960 align:middle line:84% Now let me show you one bit of evidence 00:41:37.960 --> 00:41:41.380 align:middle line:84% for the alternative approach, actually, 00:41:41.380 --> 00:41:43.300 align:middle line:84% just so that you can see why people take 00:41:43.300 --> 00:41:45.420 align:middle line:90% the other theory seriously. 00:41:45.420 --> 00:41:47.660 align:middle line:84% Here are some more sluicing examples. 00:41:47.660 --> 00:41:49.410 align:middle line:84% The classic example's from the literature. 00:41:49.410 --> 00:41:52.280 align:middle line:84% "She bought a big car, but I don't know how big." 00:41:52.280 --> 00:41:54.740 align:middle line:90% Fine. 00:41:54.740 --> 00:41:56.420 align:middle line:84% "A biography of one of the Marx brothers 00:41:56.420 --> 00:41:57.878 align:middle line:84% is going to be published this year. 00:41:57.878 --> 00:42:00.260 align:middle line:90% Guess which?" 00:42:00.260 --> 00:42:01.900 align:middle line:90% Also fine. 00:42:01.900 --> 00:42:05.670 align:middle line:84% Do people agree that these are fine? 00:42:05.670 --> 00:42:07.085 align:middle line:90% I think they're fine. 00:42:07.085 --> 00:42:08.460 align:middle line:84% All right, let's think about what 00:42:08.460 --> 00:42:13.170 align:middle line:84% would be in the part that's not pronounced. 00:42:13.170 --> 00:42:16.010 align:middle line:84% The way we've been talking, these 00:42:16.010 --> 00:42:17.600 align:middle line:84% should mean, "She bought a big car, 00:42:17.600 --> 00:42:20.250 align:middle line:84% but I don't know how big she bought a car," 00:42:20.250 --> 00:42:22.132 align:middle line:84% or "A biography of one of the Marx brothers 00:42:22.132 --> 00:42:23.590 align:middle line:84% is going to be published this year. 00:42:23.590 --> 00:42:28.080 align:middle line:84% Guess which a biography of is going to be published." 00:42:28.080 --> 00:42:30.780 align:middle line:84% And the problem with that is that these are not 00:42:30.780 --> 00:42:32.850 align:middle line:90% grammatical sentences. 00:42:32.850 --> 00:42:33.960 align:middle line:90% Do people agree with that? 00:42:33.960 --> 00:42:38.520 align:middle line:84% If you were to pronounce the whole thing, you would die. 00:42:38.520 --> 00:42:41.920 align:middle line:90% Well, maybe not die. 00:42:41.920 --> 00:42:45.580 align:middle line:84% It would depend on your previous medical condition. 00:42:45.580 --> 00:42:46.080 align:middle line:90% Yeah? 00:42:46.080 --> 00:42:49.200 align:middle line:90% AUDIENCE: [INAUDIBLE] published. 00:42:49.200 --> 00:42:51.080 align:middle line:84% NORVIN RICHARDS: I'm sorry, say it again. 00:42:51.080 --> 00:42:52.110 align:middle line:90% AUDIENCE: I meant-- 00:42:52.110 --> 00:42:53.360 align:middle line:90% NORVIN RICHARDS: You're just-- 00:42:53.360 --> 00:42:55.398 align:middle line:84% AUDIENCE: You said you would what? 00:42:55.398 --> 00:42:57.440 align:middle line:84% NORVIN RICHARDS: Oh, you're just testing my claim 00:42:57.440 --> 00:42:58.850 align:middle line:84% that if you said this aloud, you would die? 00:42:58.850 --> 00:42:59.690 align:middle line:90% Oh, gosh. 00:42:59.690 --> 00:43:00.230 align:middle line:90% So-- 00:43:00.230 --> 00:43:02.180 align:middle line:84% AUDIENCE: "A biography of one of the Marx brothers 00:43:02.180 --> 00:43:03.597 align:middle line:84% is going to be published this year 00:43:03.597 --> 00:43:05.390 align:middle line:84% and guess what the biography of (blank)." 00:43:05.390 --> 00:43:09.120 align:middle line:84% Whatever goes in the blank is going to be published. 00:43:09.120 --> 00:43:09.990 align:middle line:90% OK. 00:43:09.990 --> 00:43:12.530 align:middle line:84% NORVIN RICHARDS: How do people feel about this? 00:43:12.530 --> 00:43:14.530 align:middle line:84% I didn't say that people were going to kill you. 00:43:14.530 --> 00:43:15.760 align:middle line:90% I said you were going to die. 00:43:15.760 --> 00:43:19.020 align:middle line:84% So you shouldn't feel like you have to harm him in any way. 00:43:19.020 --> 00:43:20.752 align:middle line:90% Raquel, did-- 00:43:20.752 --> 00:43:24.310 align:middle line:84% AUDIENCE: I think the way that I think of filling it is-- 00:43:24.310 --> 00:43:27.940 align:middle line:84% "I don't know how big the car is," or "I don't know which 00:43:27.940 --> 00:43:32.500 align:middle line:84% biography is in which--" thinking [INAUDIBLE] 00:43:32.500 --> 00:43:33.640 align:middle line:90% it feels kind of weird. 00:43:33.640 --> 00:43:35.910 align:middle line:84% It feels it's weird movement going on there. 00:43:35.910 --> 00:43:36.220 align:middle line:90% NORVIN RICHARDS: Yeah. 00:43:36.220 --> 00:43:36.928 align:middle line:90% Yeah, yeah, yeah. 00:43:36.928 --> 00:43:40.460 align:middle line:90% So you seized on the move. 00:43:40.460 --> 00:43:43.340 align:middle line:84% So I said, here's a bit of evidence for the opposing camp. 00:43:43.340 --> 00:43:46.670 align:middle line:84% So the opposing camp, the people who think, no, in a sluice, 00:43:46.670 --> 00:43:48.700 align:middle line:90% there's no wh-movement going on. 00:43:48.700 --> 00:43:53.190 align:middle line:84% When you say, "I'm reading a book--" "I'm reading a book, 00:43:53.190 --> 00:43:55.040 align:middle line:90% but I can't remember what." 00:43:55.040 --> 00:43:57.385 align:middle line:84% "John is reading a book, but I don't know what." 00:43:57.385 --> 00:43:59.510 align:middle line:84% "John is reading something, but I don't know what." 00:43:59.510 --> 00:44:01.430 align:middle line:84% That sentence just ends with "what." 00:44:01.430 --> 00:44:06.003 align:middle line:84% It isn't "but I don't know what he's reading." 00:44:06.003 --> 00:44:07.670 align:middle line:84% So you don't have the complete sentence, 00:44:07.670 --> 00:44:08.960 align:middle line:84% and you're leaving part of it out. 00:44:08.960 --> 00:44:10.502 align:middle line:84% These people want to say, no, there's 00:44:10.502 --> 00:44:11.812 align:middle line:90% no wh-movement going on at all. 00:44:11.812 --> 00:44:13.520 align:middle line:84% And these kinds of examples are the kinds 00:44:13.520 --> 00:44:14.603 align:middle line:90% of examples they point at. 00:44:14.603 --> 00:44:16.992 align:middle line:84% They say, look, we know how wh-movement works. 00:44:16.992 --> 00:44:18.950 align:middle line:84% You can't do it out of certain kinds of things. 00:44:18.950 --> 00:44:20.720 align:middle line:84% We talked about in this class, that there 00:44:20.720 --> 00:44:22.580 align:middle line:84% are what are called islands, domains 00:44:22.580 --> 00:44:23.720 align:middle line:90% that you can't move out of. 00:44:23.720 --> 00:44:26.480 align:middle line:84% These are not among the islands that we talked about. 00:44:26.480 --> 00:44:28.760 align:middle line:84% I think I warned you when I showed you some islands. 00:44:28.760 --> 00:44:30.350 align:middle line:84% There are so many islands out there. 00:44:30.350 --> 00:44:31.982 align:middle line:90% It's like the Pacific out there. 00:44:31.982 --> 00:44:33.440 align:middle line:84% There are lots and lots of islands. 00:44:33.440 --> 00:44:35.390 align:middle line:84% Lots of work on charting them and figuring out 00:44:35.390 --> 00:44:38.130 align:middle line:84% how deep the water is in different places. 00:44:38.130 --> 00:44:40.558 align:middle line:84% And so here are some more islands. 00:44:40.558 --> 00:44:42.350 align:middle line:84% We don't have to worry about what they are, 00:44:42.350 --> 00:44:43.190 align:middle line:90% or why they're islands. 00:44:43.190 --> 00:44:45.482 align:middle line:84% It's clear that if you were to say this whole sentence, 00:44:45.482 --> 00:44:46.310 align:middle line:90% it would be bad. 00:44:46.310 --> 00:44:49.035 align:middle line:90% Bad things would happen to you. 00:44:49.035 --> 00:44:50.660 align:middle line:84% And the people who want to say, there's 00:44:50.660 --> 00:44:53.600 align:middle line:84% no wh-movement in a sluice, they point at stuff like this. 00:44:53.600 --> 00:44:55.940 align:middle line:84% The people who say, yes there is, too, 00:44:55.940 --> 00:44:58.468 align:middle line:84% wh-movement in a sluice, what they say 00:44:58.468 --> 00:44:59.510 align:middle line:90% is what Raquel just said. 00:44:59.510 --> 00:45:00.290 align:middle line:90% They said, no, but look. 00:45:00.290 --> 00:45:02.832 align:middle line:84% How do you know that you're not saying "She bought a big car, 00:45:02.832 --> 00:45:05.240 align:middle line:84% but I don't know how big it was." 00:45:05.240 --> 00:45:07.280 align:middle line:84% Or "A biography of one of the Marx brothers 00:45:07.280 --> 00:45:08.738 align:middle line:84% is going to be published this year. 00:45:08.738 --> 00:45:10.250 align:middle line:90% Guess which one it is"-- 00:45:10.250 --> 00:45:12.210 align:middle line:84% "...Guess which Marx brother it will be." 00:45:12.210 --> 00:45:15.060 align:middle line:90% Yeah, another possibility. 00:45:15.060 --> 00:45:18.630 align:middle line:84% Kateryna, did you have a question before? 00:45:18.630 --> 00:45:22.110 align:middle line:84% Or I'm sorry, I didn't mean to put you on track-- 00:45:22.110 --> 00:45:22.800 align:middle line:90% spot. 00:45:22.800 --> 00:45:24.990 align:middle line:84% So that's an argument that goes on between these two 00:45:24.990 --> 00:45:25.823 align:middle line:90% halves of the field. 00:45:25.823 --> 00:45:26.380 align:middle line:90% Yeah? 00:45:26.380 --> 00:45:28.122 align:middle line:84% AUDIENCE: Isn't this the essence of one 00:45:28.122 --> 00:45:29.580 align:middle line:84% of the questions that [INAUDIBLE],, 00:45:29.580 --> 00:45:37.080 align:middle line:84% and also that there might be wh- movement or there might not be, 00:45:37.080 --> 00:45:39.091 align:middle line:84% and that there are certain things that you 00:45:39.091 --> 00:45:40.433 align:middle line:90% can't move things out of? 00:45:40.433 --> 00:45:41.350 align:middle line:90% NORVIN RICHARDS: Yeah. 00:45:41.350 --> 00:45:42.040 align:middle line:90% Yeah, yeah, yeah. 00:45:42.040 --> 00:45:42.540 align:middle line:90% Yeah, no. 00:45:42.540 --> 00:45:43.210 align:middle line:90% Absolutely. 00:45:43.210 --> 00:45:47.260 align:middle line:84% So this is-- one of the things I love about linguistics is that 00:45:47.260 --> 00:45:49.970 align:middle line:84% it's very-- here we are in 24.900 and it's not all that 00:45:49.970 --> 00:45:50.470 align:middle line:90% hard. 00:45:50.470 --> 00:45:51.970 align:middle line:84% I mean, we're late in the semester. 00:45:51.970 --> 00:45:53.200 align:middle line:84% I've had to teach you a bunch of things 00:45:53.200 --> 00:45:54.350 align:middle line:90% in order to get you here. 00:45:54.350 --> 00:45:57.400 align:middle line:84% But where I've got now is standing 00:45:57.400 --> 00:46:00.130 align:middle line:84% right on the edge of the abyss looking out. 00:46:00.130 --> 00:46:03.010 align:middle line:84% So it's not clear where the bridge 00:46:03.010 --> 00:46:05.260 align:middle line:84% is, that will get us across the abyss. 00:46:05.260 --> 00:46:09.520 align:middle line:84% These are two ways that people deal with this problem. 00:46:09.520 --> 00:46:13.480 align:middle line:84% One is to say, no, there's always wh-movement. 00:46:13.480 --> 00:46:15.912 align:middle line:84% And what's going on here is that you 00:46:15.912 --> 00:46:17.120 align:middle line:90% get to rephrase these things. 00:46:17.120 --> 00:46:18.537 align:middle line:84% Notice that you have to be careful 00:46:18.537 --> 00:46:20.740 align:middle line:84% about this to avoid losing the results 00:46:20.740 --> 00:46:22.480 align:middle line:84% that we got from the previous slides. 00:46:22.480 --> 00:46:28.300 align:middle line:84% So why can't a Russian say, "He wants to talk with somebody, 00:46:28.300 --> 00:46:33.310 align:middle line:84% but I don't know who it is," leaving out the "with," 00:46:33.310 --> 00:46:34.708 align:middle line:90% which they can't do? 00:46:34.708 --> 00:46:36.250 align:middle line:84% So we have to understand what's going 00:46:36.250 --> 00:46:38.083 align:middle line:90% on in those kinds of examples. 00:46:38.083 --> 00:46:39.500 align:middle line:84% There are other things people say, 00:46:39.500 --> 00:46:42.280 align:middle line:84% but this is the place where the field currently is, 00:46:42.280 --> 00:46:44.800 align:middle line:84% trying to figure out what to do about this kind of clash 00:46:44.800 --> 00:46:46.195 align:middle line:90% between data. 00:46:46.195 --> 00:46:50.000 align:middle line:90% 00:46:50.000 --> 00:46:50.540 align:middle line:90% Cool. 00:46:50.540 --> 00:46:51.707 align:middle line:90% Questions about any of that? 00:46:51.707 --> 00:46:54.768 align:middle line:90% 00:46:54.768 --> 00:46:56.130 align:middle line:90% OK. 00:46:56.130 --> 00:47:01.900 align:middle line:84% So moral of all this, and then we will switch topics again, 00:47:01.900 --> 00:47:05.020 align:middle line:84% what we're seeing, maybe we knew this before, 00:47:05.020 --> 00:47:11.800 align:middle line:84% is that our best take on how-- so here we are, almost 00:47:11.800 --> 00:47:13.120 align:middle line:90% at the end of semantics. 00:47:13.120 --> 00:47:16.780 align:middle line:84% And what we're learning is if you 00:47:16.780 --> 00:47:19.720 align:middle line:84% are trying to interpret a sentence, what 00:47:19.720 --> 00:47:22.150 align:middle line:84% you are interpreting is not necessarily 00:47:22.150 --> 00:47:24.370 align:middle line:90% exactly what you hear. 00:47:24.370 --> 00:47:26.890 align:middle line:84% So there are phenomena like reconstruction 00:47:26.890 --> 00:47:30.460 align:middle line:84% where it sure looks like you get to take something that 00:47:30.460 --> 00:47:33.670 align:middle line:84% used to be in one place and is now in another place 00:47:33.670 --> 00:47:37.610 align:middle line:84% and interpret it as though it was still where it used to be. 00:47:37.610 --> 00:47:39.460 align:middle line:84% So "Which picture of himself did he buy?" 00:47:39.460 --> 00:47:43.253 align:middle line:84% is grammatical, because you have the option of reconstructing 00:47:43.253 --> 00:47:44.920 align:middle line:84% "Which picture of himself" to a position 00:47:44.920 --> 00:47:47.590 align:middle line:84% where he "himself" is still bound by "he." 00:47:47.590 --> 00:47:50.170 align:middle line:84% so we know that "himself" needs to be c-commanded by "he," 00:47:50.170 --> 00:47:52.570 align:middle line:84% and it isn't in the move to position. 00:47:52.570 --> 00:47:56.020 align:middle line:84% Or we think that there is such a thing as QR, quantifier 00:47:56.020 --> 00:47:58.000 align:middle line:84% raising, which creates ambiguities 00:47:58.000 --> 00:48:00.100 align:middle line:84% like the one in "Someone loves everyone." 00:48:00.100 --> 00:48:01.990 align:middle line:84% By taking "everyone" and moving it 00:48:01.990 --> 00:48:04.140 align:middle line:90% to a position above "someone." 00:48:04.140 --> 00:48:06.610 align:middle line:84% Yeah, we saw some evidence for that. 00:48:06.610 --> 00:48:08.890 align:middle line:84% In English, you can't see that, at least you 00:48:08.890 --> 00:48:11.290 align:middle line:84% don't have to say "Everyone someone 00:48:11.290 --> 00:48:13.858 align:middle line:84% loves" in order for the sentence to be ambiguous. 00:48:13.858 --> 00:48:15.400 align:middle line:84% You can say, "Someone loves everyone" 00:48:15.400 --> 00:48:16.510 align:middle line:90% and get an ambiguous sentence. 00:48:16.510 --> 00:48:18.430 align:middle line:84% We saw that there are languages, like Hungarian, where 00:48:18.430 --> 00:48:20.138 align:middle line:84% you do get to see the movement, but there 00:48:20.138 --> 00:48:23.620 align:middle line:84% are plenty of languages where you don't, like English. 00:48:23.620 --> 00:48:26.530 align:middle line:84% Or ellipsis, where we've seen some arguments, some of them 00:48:26.530 --> 00:48:28.988 align:middle line:84% pretty compelling, that when you say "She bought something, 00:48:28.988 --> 00:48:31.330 align:middle line:84% but I don't know what," that it's really "She bought 00:48:31.330 --> 00:48:33.310 align:middle line:84% something, but I don't know what she bought," 00:48:33.310 --> 00:48:35.590 align:middle line:84% and you are interpreting something that's 00:48:35.590 --> 00:48:37.210 align:middle line:90% larger than what you can hear. 00:48:37.210 --> 00:48:38.800 align:middle line:84% There's this option of leaving out 00:48:38.800 --> 00:48:42.760 align:middle line:84% some of the stuff that's syntactically there. 00:48:42.760 --> 00:48:46.985 align:middle line:84% So interpretation is complicated for all kinds of reasons. 00:48:46.985 --> 00:48:49.610 align:middle line:84% There's all kinds of interesting stuff semanticists talk about. 00:48:49.610 --> 00:48:51.370 align:middle line:84% But one of the things that they have to do 00:48:51.370 --> 00:48:54.910 align:middle line:84% is figure out what exactly is being interpreted. 00:48:54.910 --> 00:48:58.270 align:middle line:84% And it's not just a matter of figuring out what you heard. 00:48:58.270 --> 00:49:02.562 align:middle line:84% There's stuff that you didn't hear that's going on. 00:49:02.562 --> 00:49:04.645 align:middle line:84% Part of what makes semantics an interesting field. 00:49:04.645 --> 00:49:07.920 align:middle line:90% 00:49:07.920 --> 00:49:10.020 align:middle line:90% Complete shift in topic. 00:49:10.020 --> 00:49:12.930 align:middle line:84% No more talk about reconstruction, very little 00:49:12.930 --> 00:49:14.910 align:middle line:90% talk about syntax. 00:49:14.910 --> 00:49:16.780 align:middle line:84% Talk about something completely different. 00:49:16.780 --> 00:49:18.420 align:middle line:84% So if anybody would like to continue 00:49:18.420 --> 00:49:24.750 align:middle line:84% to discuss sluicing, or guards, or buildings, or ellipses, 00:49:24.750 --> 00:49:28.500 align:middle line:84% or any of this stuff, this is the time. 00:49:28.500 --> 00:49:31.142 align:middle line:84% So other kinds of things semanticists talk about. 00:49:31.142 --> 00:49:32.850 align:middle line:84% We'll see how much of this we get through 00:49:32.850 --> 00:49:34.635 align:middle line:90% before we have to stop. 00:49:34.635 --> 00:49:37.260 align:middle line:84% Consider a sentence like "I will only give Mary three cookies." 00:49:37.260 --> 00:49:40.810 align:middle line:90% 00:49:40.810 --> 00:49:44.680 align:middle line:84% That's a very ambiguous sentence. 00:49:44.680 --> 00:49:47.200 align:middle line:84% It can mean-- here's one thing it can mean. 00:49:47.200 --> 00:49:49.405 align:middle line:84% No, actually, let me say it a different way. 00:49:49.405 --> 00:49:51.530 align:middle line:84% You can say, "I will only give Mary three cookies." 00:49:51.530 --> 00:49:55.570 align:middle line:84% You can say, "I will only give Mary three cookies," which 00:49:55.570 --> 00:49:58.000 align:middle line:84% means something different from "I will only give Mary 00:49:58.000 --> 00:50:04.370 align:middle line:84% three cookies," which means something different from "I 00:50:04.370 --> 00:50:06.370 align:middle line:84% will only give Mary three cookies." 00:50:06.370 --> 00:50:09.120 align:middle line:90% Yeah. 00:50:09.120 --> 00:50:11.250 align:middle line:84% I think you can also say, "I will only 00:50:11.250 --> 00:50:13.120 align:middle line:90% give Mary three cookies." 00:50:13.120 --> 00:50:13.740 align:middle line:90% Yeah. 00:50:13.740 --> 00:50:15.810 align:middle line:90% I will not-- yeah. 00:50:15.810 --> 00:50:16.560 align:middle line:90% Yes? 00:50:16.560 --> 00:50:18.352 align:middle line:84% AUDIENCE: Another example I've seen of this 00:50:18.352 --> 00:50:20.117 align:middle line:84% is "Congratulations on your baby!" 00:50:20.117 --> 00:50:20.992 align:middle line:90% NORVIN RICHARDS: Yes? 00:50:20.992 --> 00:50:22.230 align:middle line:90% AUDIENCE: Or if you-- 00:50:22.230 --> 00:50:23.355 align:middle line:90% NORVIN RICHARDS: Thank you. 00:50:23.355 --> 00:50:25.520 align:middle line:90% 00:50:25.520 --> 00:50:27.480 align:middle line:90% AUDIENCE: If you put a focus-- 00:50:27.480 --> 00:50:30.810 align:middle line:84% every single word in the sentence except for "on," 00:50:30.810 --> 00:50:32.730 align:middle line:84% if you put your focus on that word, 00:50:32.730 --> 00:50:34.530 align:middle line:84% it means something completely different. 00:50:34.530 --> 00:50:35.655 align:middle line:90% NORVIN RICHARDS: Oh, I see. 00:50:35.655 --> 00:50:36.870 align:middle line:90% Yes. 00:50:36.870 --> 00:50:38.280 align:middle line:90% AUDIENCE: You can't say that. 00:50:38.280 --> 00:50:39.570 align:middle line:84% NORVIN RICHARDS: "Congratulations on your baby, 00:50:39.570 --> 00:50:40.500 align:middle line:90% not under your baby." 00:50:40.500 --> 00:50:41.000 align:middle line:90% Yes. 00:50:41.000 --> 00:50:43.200 align:middle line:90% Yeah, yeah, yeah. 00:50:43.200 --> 00:50:44.500 align:middle line:90% AUDIENCE: [INAUDIBLE] 00:50:44.500 --> 00:50:46.500 align:middle line:84% NORVIN RICHARDS: "Congratulations on your baby," 00:50:46.500 --> 00:50:47.640 align:middle line:90% yeah. 00:50:47.640 --> 00:50:52.200 align:middle line:84% So let's go back to the cookies, though. 00:50:52.200 --> 00:50:55.390 align:middle line:84% The phenomenon is called association with focus. 00:50:55.390 --> 00:50:57.510 align:middle line:84% And the idea is if I say something 00:50:57.510 --> 00:51:02.950 align:middle line:84% like "I will only give Mary three cookies," 00:51:02.950 --> 00:51:06.850 align:middle line:84% what I'm doing is inviting you to consider 00:51:06.850 --> 00:51:09.160 align:middle line:84% all of the sentences of the form, 00:51:09.160 --> 00:51:12.810 align:middle line:90% "I will give x three cookies." 00:51:12.810 --> 00:51:15.890 align:middle line:84% And I'm asserting all of these are false, 00:51:15.890 --> 00:51:18.980 align:middle line:84% except for the one where x is Mary. 00:51:18.980 --> 00:51:20.010 align:middle line:90% That's what that means. 00:51:20.010 --> 00:51:22.160 align:middle line:84% So "I will only give Mary three cookies" 00:51:22.160 --> 00:51:24.200 align:middle line:84% means I will give Mary three cookies, 00:51:24.200 --> 00:51:27.590 align:middle line:84% and I will not give three cookies to Bill, or John, 00:51:27.590 --> 00:51:31.410 align:middle line:84% or Susan, or Fred, or anybody else that's relevant. 00:51:31.410 --> 00:51:36.590 align:middle line:84% Similarly, if I say "I will only give Mary three cookies," 00:51:36.590 --> 00:51:38.870 align:middle line:84% it means I will give her three cookies, 00:51:38.870 --> 00:51:42.380 align:middle line:84% but I will not give her three hamburgers, or three hot dogs, 00:51:42.380 --> 00:51:43.670 align:middle line:90% or three cakes. 00:51:43.670 --> 00:51:45.740 align:middle line:90% Only three cookies. 00:51:45.740 --> 00:51:50.330 align:middle line:84% So association with focus, it's this interesting phenomenon 00:51:50.330 --> 00:51:53.450 align:middle line:84% that shows up where, depending on which of these words 00:51:53.450 --> 00:51:57.950 align:middle line:84% I'm saying loudest, what I'm inviting you to do 00:51:57.950 --> 00:52:01.700 align:middle line:84% is consider other sentences where that word has been 00:52:01.700 --> 00:52:03.155 align:middle line:90% substituted for something else. 00:52:03.155 --> 00:52:07.430 align:middle line:90% 00:52:07.430 --> 00:52:09.290 align:middle line:84% And in this particular case, what 00:52:09.290 --> 00:52:11.790 align:middle line:84% I'm asserting is that all the other sentences are false, 00:52:11.790 --> 00:52:15.920 align:middle line:84% all the ones except the one that I've said. 00:52:15.920 --> 00:52:19.500 align:middle line:84% People generally decompose this into two parts. 00:52:19.500 --> 00:52:23.750 align:middle line:84% There's focus, which is the putting special emphasis 00:52:23.750 --> 00:52:26.210 align:middle line:84% on a particular word thing, which 00:52:26.210 --> 00:52:29.373 align:middle line:84% invites you to consider all of the alternatives to that word. 00:52:29.373 --> 00:52:31.790 align:middle line:84% And what you're supposed to do with the other alternatives 00:52:31.790 --> 00:52:33.350 align:middle line:84% to that word depends on what goes on 00:52:33.350 --> 00:52:34.880 align:middle line:90% in the rest of the sentence. 00:52:34.880 --> 00:52:38.940 align:middle line:84% So "only" associates with focus to say, 00:52:38.940 --> 00:52:40.970 align:middle line:84% all the other alternatives are false, 00:52:40.970 --> 00:52:42.330 align:middle line:90% give you false sentences. 00:52:42.330 --> 00:52:44.270 align:middle line:84% So "I will only give Mary three cookies" 00:52:44.270 --> 00:52:46.130 align:middle line:84% means I will give Mary three cookies, 00:52:46.130 --> 00:52:48.470 align:middle line:84% but I won't give three cookies to anyone else. 00:52:48.470 --> 00:52:53.930 align:middle line:84% If I say, "I will even give Mary three cookies," that 00:52:53.930 --> 00:52:57.140 align:middle line:84% means something like I will give Mary three cookies, 00:52:57.140 --> 00:52:59.360 align:middle line:84% and she is the least likely person for me 00:52:59.360 --> 00:53:01.190 align:middle line:90% to give three cookies. 00:53:01.190 --> 00:53:03.590 align:middle line:90% We all know that I hate Mary. 00:53:03.590 --> 00:53:05.900 align:middle line:84% But I'm going to give three cookies to everybody, 00:53:05.900 --> 00:53:09.380 align:middle line:84% even Mary, who I hate, and also she's diabetic. 00:53:09.380 --> 00:53:11.660 align:middle line:84% But I'm going to give her three cookies anyway. 00:53:11.660 --> 00:53:14.213 align:middle line:90% Yeah? 00:53:14.213 --> 00:53:15.880 align:middle line:84% Or "I will even give Mary three cookies" 00:53:15.880 --> 00:53:20.530 align:middle line:84% means I will give her three cookies, not just 00:53:20.530 --> 00:53:23.200 align:middle line:90% other numbers, like two or one. 00:53:23.200 --> 00:53:25.450 align:middle line:90% That's what that means. 00:53:25.450 --> 00:53:28.030 align:middle line:84% "I will give Mary three cookies, too," 00:53:28.030 --> 00:53:30.520 align:middle line:84% actually says, I will give Mary three cookies, 00:53:30.520 --> 00:53:32.830 align:middle line:84% and there's at least one alternative, which 00:53:32.830 --> 00:53:34.338 align:middle line:90% is also true, I guess. 00:53:34.338 --> 00:53:36.130 align:middle line:84% So "I will give Mary three cookies" means-- 00:53:36.130 --> 00:53:37.450 align:middle line:84% "I will give Mary three cookies, too," 00:53:37.450 --> 00:53:39.430 align:middle line:84% means I'll give Mary three cookies in addition 00:53:39.430 --> 00:53:41.260 align:middle line:84% to the other person that I'm going to give three cookies to, 00:53:41.260 --> 00:53:42.790 align:middle line:84% which maybe you already know about. 00:53:42.790 --> 00:53:43.397 align:middle line:90% Yeah? 00:53:43.397 --> 00:53:45.730 align:middle line:84% AUDIENCE: In the middle one, if you emphasize "cookies," 00:53:45.730 --> 00:53:47.860 align:middle line:84% it would be like giving her three of other things, 00:53:47.860 --> 00:53:49.780 align:middle line:84% but you're also throwing in three cookies. 00:53:49.780 --> 00:53:50.697 align:middle line:90% NORVIN RICHARDS: Yeah. 00:53:50.697 --> 00:53:53.560 align:middle line:84% So "I will give Mary three cookies, too." 00:53:53.560 --> 00:53:55.660 align:middle line:84% So this is another interesting aspect 00:53:55.660 --> 00:53:56.950 align:middle line:90% of association with focus. 00:53:56.950 --> 00:54:01.130 align:middle line:90% 00:54:01.130 --> 00:54:04.242 align:middle line:84% Yes, so it can mean what you just said. 00:54:04.242 --> 00:54:05.450 align:middle line:90% I'm going to write that down. 00:54:05.450 --> 00:54:08.312 align:middle line:90% 00:54:08.312 --> 00:54:18.500 align:middle line:84% "I will give Mary three cookies, too." 00:54:18.500 --> 00:54:20.780 align:middle line:84% It can mean, I will give her three cookies 00:54:20.780 --> 00:54:23.360 align:middle line:84% and I will also give her three of something else. 00:54:23.360 --> 00:54:25.543 align:middle line:90% You're absolutely right. 00:54:25.543 --> 00:54:27.710 align:middle line:84% I feel as though it can mean something else as well. 00:54:27.710 --> 00:54:31.190 align:middle line:84% "I will give Mary three cookies, too." 00:54:31.190 --> 00:54:34.120 align:middle line:84% Can people get that to mean something else? 00:54:34.120 --> 00:54:35.650 align:middle line:84% I feel as though it can also mean, 00:54:35.650 --> 00:54:37.570 align:middle line:84% not only will I give her three cookies, 00:54:37.570 --> 00:54:40.373 align:middle line:90% but I will give her a bicycle. 00:54:40.373 --> 00:54:42.290 align:middle line:84% That is, the other thing that I will give her, 00:54:42.290 --> 00:54:44.170 align:middle line:90% it doesn't have to be three. 00:54:44.170 --> 00:54:49.150 align:middle line:84% So I will give Mary a cake, I will give Mary a pie, 00:54:49.150 --> 00:54:52.520 align:middle line:84% and I will give her three cookies, too. 00:54:52.520 --> 00:54:53.520 align:middle line:90% Yeah. 00:54:53.520 --> 00:54:57.200 align:middle line:84% So this is an interesting aspect of all this 00:54:57.200 --> 00:54:58.580 align:middle line:90% that people work on a lot. 00:54:58.580 --> 00:55:02.570 align:middle line:84% So I've just been talking about putting stress on the word 00:55:02.570 --> 00:55:04.110 align:middle line:84% and saying you get to consider all 00:55:04.110 --> 00:55:05.360 align:middle line:90% the alternatives to that word. 00:55:05.360 --> 00:55:07.730 align:middle line:84% And that is true, that if you put stress on a word, 00:55:07.730 --> 00:55:09.980 align:middle line:84% you get to consider all the alternatives to this word. 00:55:09.980 --> 00:55:12.140 align:middle line:84% There's this phenomenon is sometimes 00:55:12.140 --> 00:55:13.700 align:middle line:90% called focus projection. 00:55:13.700 --> 00:55:15.810 align:middle line:84% And one way to think about it is, 00:55:15.810 --> 00:55:18.650 align:middle line:84% we know how to focus a word, you say the word louder. 00:55:18.650 --> 00:55:21.260 align:middle line:90% How do you focus a phrase? 00:55:21.260 --> 00:55:24.440 align:middle line:84% And maybe what we're learning is a way 00:55:24.440 --> 00:55:27.490 align:middle line:84% to focus the whole phrase, "three cookies," 00:55:27.490 --> 00:55:30.010 align:middle line:84% is the same way you would focus "cookies." 00:55:30.010 --> 00:55:31.870 align:middle line:84% That is, you put stress on "cookies." 00:55:31.870 --> 00:55:34.078 align:middle line:84% You might have thought that the way to focus a phrase 00:55:34.078 --> 00:55:35.950 align:middle line:84% would be to say the whole phrase louder. 00:55:35.950 --> 00:55:39.250 align:middle line:84% So "I will give Mary a cake, I will give her a pie, 00:55:39.250 --> 00:55:41.950 align:middle line:84% and I will give her three cookies, too." 00:55:41.950 --> 00:55:44.830 align:middle line:84% You might have thought that you would have to do that. 00:55:44.830 --> 00:55:45.460 align:middle line:90% You don't. 00:55:45.460 --> 00:55:50.380 align:middle line:84% You can, maybe, but you don't have to do that. 00:55:50.380 --> 00:55:54.610 align:middle line:84% One of your options is to just emphasize 00:55:54.610 --> 00:55:57.580 align:middle line:84% part of the phrase, and now the theory, 00:55:57.580 --> 00:56:00.020 align:middle line:84% or the literature is off and running: 00:56:00.020 --> 00:56:03.070 align:middle line:84% which part of the phrase do you get to emphasize in order 00:56:03.070 --> 00:56:04.630 align:middle line:90% to focus the whole phrase? 00:56:04.630 --> 00:56:07.150 align:middle line:84% Because notice there are constraints. 00:56:07.150 --> 00:56:14.410 align:middle line:84% If I say, "I will give Mary three cookies, too," well. 00:56:14.410 --> 00:56:19.640 align:middle line:84% That's grammatical, but the only alternatives are other numbers. 00:56:19.640 --> 00:56:21.353 align:middle line:84% So "I will give Mary three cookies, too," 00:56:21.353 --> 00:56:23.770 align:middle line:84% means not only will I give her one cookie and two cookies, 00:56:23.770 --> 00:56:25.420 align:middle line:84% but I will also give her three cookies. 00:56:25.420 --> 00:56:26.320 align:middle line:90% It can mean that. 00:56:26.320 --> 00:56:28.930 align:middle line:84% But "I will give her a cake, and I will give her a pie, 00:56:28.930 --> 00:56:32.230 align:middle line:84% and I will also give her three cookies"-- 00:56:32.230 --> 00:56:32.980 align:middle line:90% that's peculiar. 00:56:32.980 --> 00:56:35.870 align:middle line:90% 00:56:35.870 --> 00:56:38.810 align:middle line:84% If I want to avoid being peculiar, 00:56:38.810 --> 00:56:41.720 align:middle line:90% I have a long way to go. 00:56:41.720 --> 00:56:43.770 align:middle line:90% But I can start. 00:56:43.770 --> 00:56:46.160 align:middle line:84% The first step on the long, long journey 00:56:46.160 --> 00:56:47.960 align:middle line:84% is to avoid saying things like that. 00:56:47.960 --> 00:56:50.150 align:middle line:90% Instead, I need to say-- 00:56:50.150 --> 00:56:53.055 align:middle line:90% sorry, I'm cracking myself up. 00:56:53.055 --> 00:56:54.680 align:middle line:84% What I need to say instead is something 00:56:54.680 --> 00:56:57.560 align:middle line:84% like "I will give her a cake, and I will give her a pie, 00:56:57.560 --> 00:56:59.320 align:middle line:84% and I will give her three cookies, too"-- 00:56:59.320 --> 00:57:00.830 align:middle line:90% to put the stress on "cookies." 00:57:00.830 --> 00:57:02.995 align:middle line:84% That's the normal way to focus the whole phrase 00:57:02.995 --> 00:57:05.120 align:middle line:84% "three cookies," is to put the stress on "cookies." 00:57:05.120 --> 00:57:08.660 align:middle line:84% So how come it's on the "cookies" and not on "three"? 00:57:08.660 --> 00:57:11.750 align:middle line:84% Maybe because the whole phrase is a noun phrase and "cookies" 00:57:11.750 --> 00:57:13.982 align:middle line:90% is the noun. 00:57:13.982 --> 00:57:15.690 align:middle line:84% Or maybe it's more complicated than that. 00:57:15.690 --> 00:57:19.348 align:middle line:84% There's all kinds of work now to figure out what's going on. 00:57:19.348 --> 00:57:21.140 align:middle line:84% So there's a phenomenon in focus projection 00:57:21.140 --> 00:57:23.515 align:middle line:84% which people get very excited about trying to figure out. 00:57:23.515 --> 00:57:26.360 align:middle line:84% What is it that lets you do this? 00:57:26.360 --> 00:57:28.370 align:middle line:84% Because it goes further than this-- 00:57:28.370 --> 00:57:33.530 align:middle line:84% So actually, let me destroy the theory that I just offered. 00:57:33.530 --> 00:57:48.630 align:middle line:84% Consider a sentence like "I will only talk about bats." 00:57:48.630 --> 00:57:52.600 align:middle line:90% 00:57:52.600 --> 00:57:55.360 align:middle line:84% Now imagine that the main thing that I emphasize is "bats." 00:57:55.360 --> 00:57:58.010 align:middle line:90% 00:57:58.010 --> 00:58:02.240 align:middle line:84% So that has one meaning, that's the easy meaning. 00:58:02.240 --> 00:58:05.360 align:middle line:84% "I will only talk about bats" means I will not 00:58:05.360 --> 00:58:07.058 align:middle line:90% talk about anything else. 00:58:07.058 --> 00:58:08.225 align:middle line:90% I will only talk about bats. 00:58:08.225 --> 00:58:10.693 align:middle line:90% 00:58:10.693 --> 00:58:12.860 align:middle line:84% But I think there are some other things it can mean. 00:58:12.860 --> 00:58:14.990 align:middle line:84% One of them, for example, is I can say 00:58:14.990 --> 00:58:16.790 align:middle line:90% "I will only talk about bats." 00:58:16.790 --> 00:58:20.180 align:middle line:84% I will not draw any disturbing drawings, 00:58:20.180 --> 00:58:22.610 align:middle line:90% or sing any weird songs. 00:58:22.610 --> 00:58:25.220 align:middle line:84% All I will do is talk about bats. 00:58:25.220 --> 00:58:27.110 align:middle line:90% I think it can mean that. 00:58:27.110 --> 00:58:30.470 align:middle line:84% So I can be contrasting the bats with other things, 00:58:30.470 --> 00:58:32.930 align:middle line:84% but I can also be contrasting the whole verb phrase, 00:58:32.930 --> 00:58:37.490 align:middle line:84% "talk about bats," with other verb phrases 00:58:37.490 --> 00:58:38.450 align:middle line:90% that I could be doing. 00:58:38.450 --> 00:58:38.950 align:middle line:90% Yeah? 00:58:38.950 --> 00:58:40.867 align:middle line:84% AUDIENCE: I feel like if you get that meaning, 00:58:40.867 --> 00:58:42.410 align:middle line:84% you would have to emphasize talk. 00:58:42.410 --> 00:58:44.640 align:middle line:90% "I will only talk about them. 00:58:44.640 --> 00:58:46.810 align:middle line:84% I won't sing about them, I won't draw them." 00:58:46.810 --> 00:58:48.660 align:middle line:84% NORVIN RICHARDS: Ah, that's a nice example. 00:58:48.660 --> 00:58:50.352 align:middle line:90% Let's do that. 00:58:50.352 --> 00:58:51.790 align:middle line:90% "I will only talk about bats." 00:58:51.790 --> 00:58:58.720 align:middle line:90% 00:58:58.720 --> 00:59:00.160 align:middle line:84% I think Faith is absolutely right. 00:59:00.160 --> 00:59:03.942 align:middle line:84% It's possible to say "I will only talk about bats." 00:59:03.942 --> 00:59:05.650 align:middle line:84% And I think the alternatives are the ones 00:59:05.650 --> 00:59:06.608 align:middle line:90% that you just outlined. 00:59:06.608 --> 00:59:08.950 align:middle line:84% It means I won't sing about bats, 00:59:08.950 --> 00:59:10.777 align:middle line:84% I won't compose poetry about bats, 00:59:10.777 --> 00:59:11.860 align:middle line:90% I won't dress up as a bat. 00:59:11.860 --> 00:59:13.540 align:middle line:84% All I will do is talk about bats. 00:59:13.540 --> 00:59:16.440 align:middle line:90% Yeah, I think that's right. 00:59:16.440 --> 00:59:20.400 align:middle line:90% But notice-- I think this one-- 00:59:20.400 --> 00:59:24.735 align:middle line:84% imagine that what I want to say is "I'm obsessed with bats, 00:59:24.735 --> 00:59:25.235 align:middle line:90% I'm sorry. 00:59:25.235 --> 00:59:28.790 align:middle line:90% 00:59:28.790 --> 00:59:30.950 align:middle line:90% And in fact, I-- 00:59:30.950 --> 00:59:34.440 align:middle line:84% but I'm going to do my best to behave normally at the party. 00:59:34.440 --> 00:59:39.200 align:middle line:84% So I'm not going to try to swim in the punch bowl, 00:59:39.200 --> 00:59:44.030 align:middle line:84% and I'm not going to scare the host's pets, 00:59:44.030 --> 00:59:47.180 align:middle line:84% and I'm not going to eat all the food. 00:59:47.180 --> 00:59:49.908 align:middle line:84% I'm only going to talk about bats, which is a little weird. 00:59:49.908 --> 00:59:52.450 align:middle line:84% But it's not as weird as some of the things that I could do." 00:59:52.450 --> 00:59:55.660 align:middle line:90% 00:59:55.660 --> 00:59:57.730 align:middle line:84% So if we're considering a bunch of verb phrases 00:59:57.730 --> 01:00:00.780 align:middle line:84% that have nothing to do with each other, 01:00:00.780 --> 01:00:04.700 align:middle line:84% I think one way to do it is by putting the emphasis on "bats," 01:00:04.700 --> 01:00:07.620 align:middle line:84% to say "I will only talk about bats." 01:00:07.620 --> 01:00:12.310 align:middle line:84% ("I won't try to eat the host's dog. 01:00:12.310 --> 01:00:13.420 align:middle line:90% I'll just talk about bats. 01:00:13.420 --> 01:00:15.160 align:middle line:90% Is that so bad? 01:00:15.160 --> 01:00:18.640 align:middle line:90% I like bats.") Yeah? 01:00:18.640 --> 01:00:20.170 align:middle line:90% I think this is true. 01:00:20.170 --> 01:00:24.600 align:middle line:84% Has everybody seen the video about the bats on the swing? 01:00:24.600 --> 01:00:25.203 align:middle line:90% Oh, OK. 01:00:25.203 --> 01:00:27.120 align:middle line:84% I'll try to find it and put it on the website. 01:00:27.120 --> 01:00:28.290 align:middle line:84% It has nothing to do with linguistics, 01:00:28.290 --> 01:00:29.373 align:middle line:90% but it's an awesome video. 01:00:29.373 --> 01:00:32.770 align:middle line:90% 01:00:32.770 --> 01:00:33.910 align:middle line:90% So it looks as though-- 01:00:33.910 --> 01:00:35.710 align:middle line:84% I think Faith is exactly right that if you 01:00:35.710 --> 01:00:40.840 align:middle line:84% put the stress on "talk," the alternatives are talk-- 01:00:40.840 --> 01:00:42.050 align:middle line:90% are to the verb. 01:00:42.050 --> 01:00:43.360 align:middle line:90% So I'll only talk about bats. 01:00:43.360 --> 01:00:45.735 align:middle line:84% I won't sing about bats, I won't do an interpretive dance 01:00:45.735 --> 01:00:47.470 align:middle line:90% about bats. 01:00:47.470 --> 01:00:49.293 align:middle line:84% But if I put stress on "bats," then we 01:00:49.293 --> 01:00:50.710 align:middle line:84% can consider alternatives to bats, 01:00:50.710 --> 01:00:52.252 align:middle line:84% but we can also consider alternatives 01:00:52.252 --> 01:00:53.797 align:middle line:90% to the whole verb phrase. 01:00:53.797 --> 01:00:55.630 align:middle line:84% And so this is the kind of thing, as I said, 01:00:55.630 --> 01:00:56.980 align:middle line:84% people who work on focus projection 01:00:56.980 --> 01:00:58.438 align:middle line:84% try to figure out what the heck are 01:00:58.438 --> 01:01:02.000 align:middle line:84% the rules for focus projection, because they're complicated. 01:01:02.000 --> 01:01:02.500 align:middle line:90% Yeah? 01:01:02.500 --> 01:01:03.917 align:middle line:84% AUDIENCE: Oh, it's just a question 01:01:03.917 --> 01:01:06.130 align:middle line:84% because I feel like my ability to interpret meanings 01:01:06.130 --> 01:01:06.760 align:middle line:90% has gone out the window. 01:01:06.760 --> 01:01:07.195 align:middle line:90% NORVIN RICHARDS: I understand. 01:01:07.195 --> 01:01:07.961 align:middle line:90% [LAUGHTER] 01:01:07.961 --> 01:01:11.645 align:middle line:84% AUDIENCE: What happens if you put the emphasis on "only?" 01:01:11.645 --> 01:01:12.770 align:middle line:90% NORVIN RICHARDS: On "only?" 01:01:12.770 --> 01:01:14.150 align:middle line:90% I will only talk about bats. 01:01:14.150 --> 01:01:16.780 align:middle line:90% 01:01:16.780 --> 01:01:18.197 align:middle line:84% I don't know, what does that mean? 01:01:18.197 --> 01:01:19.655 align:middle line:84% AUDIENCE: I won't do anything else. 01:01:19.655 --> 01:01:21.490 align:middle line:84% NORVIN RICHARDS: I won't do anything else. 01:01:21.490 --> 01:01:23.910 align:middle line:90% I will only talk about bats. 01:01:23.910 --> 01:01:26.160 align:middle line:84% AUDIENCE: I feel like it could call into question 01:01:26.160 --> 01:01:28.650 align:middle line:90% everything in the little brace. 01:01:28.650 --> 01:01:30.840 align:middle line:84% So you say "I will only talk about bats," 01:01:30.840 --> 01:01:32.910 align:middle line:84% and that could mean that you're saying 01:01:32.910 --> 01:01:35.370 align:middle line:84% as opposed to anything else about bats, anything 01:01:35.370 --> 01:01:36.090 align:middle line:90% else about-- 01:01:36.090 --> 01:01:37.483 align:middle line:90% or talk about any other-- 01:01:37.483 --> 01:01:38.400 align:middle line:90% NORVIN RICHARDS: Yeah. 01:01:38.400 --> 01:01:39.210 align:middle line:90% Yeah. 01:01:39.210 --> 01:01:40.530 align:middle line:90% Yeah. 01:01:40.530 --> 01:01:42.490 align:middle line:90% I think that might be true. 01:01:42.490 --> 01:01:44.465 align:middle line:90% I think that might be right. 01:01:44.465 --> 01:01:44.965 align:middle line:90% Yeah. 01:01:44.965 --> 01:01:49.190 align:middle line:90% 01:01:49.190 --> 01:01:50.060 align:middle line:90% More mysteries. 01:01:50.060 --> 01:01:52.970 align:middle line:90% 01:01:52.970 --> 01:01:57.723 align:middle line:84% Any other questions about bats, or "only," or focus projection? 01:01:57.723 --> 01:01:59.390 align:middle line:84% It's just a phenomenon that's out there. 01:01:59.390 --> 01:01:59.890 align:middle line:90% Yeah? 01:01:59.890 --> 01:02:03.290 align:middle line:84% AUDIENCE: Have we gotten across a sentence where 01:02:03.290 --> 01:02:06.020 align:middle line:84% stressing a particular word doesn't actually 01:02:06.020 --> 01:02:07.066 align:middle line:90% mean anything different? 01:02:07.066 --> 01:02:11.435 align:middle line:84% Because I feel you could stress any of the words in "I 01:02:11.435 --> 01:02:12.560 align:middle line:90% will only talk about bats." 01:02:12.560 --> 01:02:13.962 align:middle line:90% I will only talk about bats. 01:02:13.962 --> 01:02:15.170 align:middle line:90% NORVIN RICHARDS: That's true. 01:02:15.170 --> 01:02:16.180 align:middle line:84% So we've only talked about things 01:02:16.180 --> 01:02:17.380 align:middle line:90% that come after the "only." 01:02:17.380 --> 01:02:19.585 align:middle line:90% So I will only talk about bats. 01:02:19.585 --> 01:02:22.774 align:middle line:90% 01:02:22.774 --> 01:02:24.150 align:middle line:90% AUDIENCE: And not you or-- 01:02:24.150 --> 01:02:26.422 align:middle line:90% 01:02:26.422 --> 01:02:28.630 align:middle line:84% NORVIN RICHARDS: OK, first, let's forget about "only" 01:02:28.630 --> 01:02:30.770 align:middle line:90% for a second. 01:02:30.770 --> 01:02:34.850 align:middle line:84% If I say, I will talk about that, remember focus. 01:02:34.850 --> 01:02:37.180 align:middle line:84% So there's focus, which says let's consider 01:02:37.180 --> 01:02:39.850 align:middle line:84% all the alternatives to the thing that's focused. 01:02:39.850 --> 01:02:42.070 align:middle line:84% And now since we've talked about focus projection, 01:02:42.070 --> 01:02:44.710 align:middle line:84% we're bearing in mind the fact that what's focused 01:02:44.710 --> 01:02:47.020 align:middle line:90% is not necessarily what's loud. 01:02:47.020 --> 01:02:51.070 align:middle line:84% There's what's loud, which is contained in what's focused, 01:02:51.070 --> 01:02:52.780 align:middle line:84% and it might just be what's focused. 01:02:52.780 --> 01:02:54.520 align:middle line:84% But the thing that's loud could just 01:02:54.520 --> 01:02:57.040 align:middle line:84% be one part of the whole phrase that's focused. 01:02:57.040 --> 01:02:59.730 align:middle line:90% 01:02:59.730 --> 01:03:01.160 align:middle line:84% So forget about only for a second. 01:03:01.160 --> 01:03:08.030 align:middle line:84% If I say "I will talk about that," there's focus on "I," 01:03:08.030 --> 01:03:10.208 align:middle line:84% and I think I'm inviting-- out of the blue, 01:03:10.208 --> 01:03:11.750 align:middle line:84% I think I'm inviting you to entertain 01:03:11.750 --> 01:03:14.760 align:middle line:84% the possibility that all the alternatives are false. 01:03:14.760 --> 01:03:17.390 align:middle line:84% So if I say "I will talk about bats," 01:03:17.390 --> 01:03:19.518 align:middle line:84% I mean "I will talk about bats and for some reason, 01:03:19.518 --> 01:03:21.560 align:middle line:84% the idea that somebody else might talk about bats 01:03:21.560 --> 01:03:25.590 align:middle line:84% is salient and it will not happen." 01:03:25.590 --> 01:03:32.240 align:middle line:84% So for example, if you ask me "Is Faith 01:03:32.240 --> 01:03:34.550 align:middle line:84% going to talk about bats at the party?" 01:03:34.550 --> 01:03:37.640 align:middle line:84% I can say, "No, I will talk about bats," 01:03:37.640 --> 01:03:39.230 align:middle line:84% and that means I will talk about bats 01:03:39.230 --> 01:03:42.600 align:middle line:84% and the salient alternative gives you a false sentence. 01:03:42.600 --> 01:03:45.000 align:middle line:90% That's what that means. 01:03:45.000 --> 01:03:47.750 align:middle line:84% So I think when we say "I will only 01:03:47.750 --> 01:03:55.430 align:middle line:84% talk about bats," that's similarly something you can 01:03:55.430 --> 01:03:56.970 align:middle line:90% use in that kind of situation. 01:03:56.970 --> 01:04:01.760 align:middle line:84% So if you ask me, "Will Enrico only 01:04:01.760 --> 01:04:04.400 align:middle line:90% talk about bats at the party?" 01:04:04.400 --> 01:04:05.570 align:middle line:90% I say, "No, no. 01:04:05.570 --> 01:04:07.760 align:middle line:90% I will only talk about bats." 01:04:07.760 --> 01:04:11.348 align:middle line:84% And that means I am the person such that the only thing I'll 01:04:11.348 --> 01:04:12.140 align:middle line:90% talk about is bats. 01:04:12.140 --> 01:04:15.860 align:middle line:84% Enrico is more normal than me, like most people. 01:04:15.860 --> 01:04:16.580 align:middle line:90% Yeah, Kateryna? 01:04:16.580 --> 01:04:18.163 align:middle line:84% AUDIENCE: What if the more natural way 01:04:18.163 --> 01:04:20.385 align:middle line:84% to say that sentence is "only I?" 01:04:20.385 --> 01:04:22.010 align:middle line:84% NORVIN RICHARDS: So what's complicating 01:04:22.010 --> 01:04:23.385 align:middle line:84% this is that it's possible to get 01:04:23.385 --> 01:04:26.230 align:middle line:84% a meaning that's sort of like "only," without an "only." 01:04:26.230 --> 01:04:29.000 align:middle line:84% that's why I started with "I will talk about bats," 01:04:29.000 --> 01:04:32.420 align:middle line:84% which means I will talk about bats and the alternatives to me 01:04:32.420 --> 01:04:34.220 align:middle line:90% won't. 01:04:34.220 --> 01:04:36.210 align:middle line:84% Maybe if we do one of these other ones. 01:04:36.210 --> 01:04:41.640 align:middle line:84% So "I will even talk about bats" means several things. 01:04:41.640 --> 01:04:43.670 align:middle line:84% It can mean "I will talk about bats in addition 01:04:43.670 --> 01:04:47.180 align:middle line:84% to the many other things I will talk about." 01:04:47.180 --> 01:04:52.670 align:middle line:84% It can also mean "You wanted somebody to entertain everybody 01:04:52.670 --> 01:04:53.610 align:middle line:90% at the party. 01:04:53.610 --> 01:04:57.440 align:middle line:84% I will juggle, I will dance, I will even talk about bats. 01:04:57.440 --> 01:04:59.480 align:middle line:84% So I will do all the many talented things 01:04:59.480 --> 01:05:01.220 align:middle line:90% that I'm capable of doing. 01:05:01.220 --> 01:05:02.450 align:middle line:90% People will be agog. 01:05:02.450 --> 01:05:04.530 align:middle line:90% You'll be glad you invited me." 01:05:04.530 --> 01:05:06.240 align:middle line:90% It can mean that. 01:05:06.240 --> 01:05:08.720 align:middle line:84% So putting emphasis on "bats" like that 01:05:08.720 --> 01:05:11.840 align:middle line:84% can either invite you to consider the alternatives 01:05:11.840 --> 01:05:15.170 align:middle line:84% to bats, or the alternatives to talk about bats. 01:05:15.170 --> 01:05:16.610 align:middle line:90% Yeah. 01:05:16.610 --> 01:05:21.640 align:middle line:84% "I will even talk about bats," I think, 01:05:21.640 --> 01:05:23.338 align:middle line:90% can also mean that, right? 01:05:23.338 --> 01:05:24.130 align:middle line:90% That kind of thing. 01:05:24.130 --> 01:05:26.380 align:middle line:84% "I will even talk about bats" can 01:05:26.380 --> 01:05:31.440 align:middle line:84% mean "I will talk about bats, and so will 01:05:31.440 --> 01:05:34.560 align:middle line:90% all of the linguistics TAs. 01:05:34.560 --> 01:05:37.500 align:middle line:84% Various people will talk about bats, and I'm one." 01:05:37.500 --> 01:05:38.820 align:middle line:90% I will even talk about bats. 01:05:38.820 --> 01:05:40.690 align:middle line:84% John will talk about bats, Mary will talk about bats, 01:05:40.690 --> 01:05:41.857 align:middle line:90% I will even talk about bats. 01:05:41.857 --> 01:05:43.780 align:middle line:90% I think it can mean that. 01:05:43.780 --> 01:05:44.280 align:middle line:90% Faith? 01:05:44.280 --> 01:05:46.590 align:middle line:84% AUDIENCE: Aren't there even further divisions 01:05:46.590 --> 01:05:49.980 align:middle line:84% of variations of tone where, like, if you were to say 01:05:49.980 --> 01:05:53.610 align:middle line:90% "I will talk about bats." 01:05:53.610 --> 01:05:56.910 align:middle line:84% You can't say that if someone were like, oh yeah, 01:05:56.910 --> 01:05:58.990 align:middle line:84% you're going to talk about bats at this party. 01:05:58.990 --> 01:06:01.300 align:middle line:84% It just doesn't sound like an appropriate [INAUDIBLE].. 01:06:01.300 --> 01:06:03.050 align:middle line:84% NORVIN RICHARDS: "I will talk about bats." 01:06:03.050 --> 01:06:05.230 align:middle line:90% 01:06:05.230 --> 01:06:07.650 align:middle line:90% Yeah, so we actually-- 01:06:07.650 --> 01:06:09.400 align:middle line:84% did we touch on something related to this? 01:06:09.400 --> 01:06:11.560 align:middle line:90% I think we did. 01:06:11.560 --> 01:06:13.810 align:middle line:84% Yes, we did, because we were talking about pragmatics, 01:06:13.810 --> 01:06:15.370 align:middle line:84% and I was trying to get you to stop. 01:06:15.370 --> 01:06:19.420 align:middle line:84% Not you specifically, but the whole class en masse. 01:06:19.420 --> 01:06:21.850 align:middle line:84% You were all trying to force me to talk about pragmatics, 01:06:21.850 --> 01:06:22.900 align:middle line:84% and I was trying to stop you, and I 01:06:22.900 --> 01:06:25.108 align:middle line:84% don't know why I'm bringing this up now because we're 01:06:25.108 --> 01:06:27.170 align:middle line:90% about to do it again. 01:06:27.170 --> 01:06:31.960 align:middle line:84% And I was giving examples like if you ask me "How did people 01:06:31.960 --> 01:06:33.010 align:middle line:90% do on the test?" 01:06:33.010 --> 01:06:37.000 align:middle line:90% I can say, "Well, Mary passed." 01:06:37.000 --> 01:06:39.220 align:middle line:90% And then you get-- 01:06:39.220 --> 01:06:43.330 align:middle line:84% that means-- all I've told you is that Mary passed. 01:06:43.330 --> 01:06:46.240 align:middle line:84% And if we were all robots, you might 01:06:46.240 --> 01:06:48.340 align:middle line:84% have expected that all you would conclude 01:06:48.340 --> 01:06:50.230 align:middle line:90% from that is that Mary passed. 01:06:50.230 --> 01:06:52.180 align:middle line:84% So you start that conversation not knowing 01:06:52.180 --> 01:06:53.480 align:middle line:90% how people did on the test. 01:06:53.480 --> 01:06:55.210 align:middle line:84% That's why you asked me, "How did people do on the test?" 01:06:55.210 --> 01:06:56.818 align:middle line:84% I told you Mary passed, and now you 01:06:56.818 --> 01:06:59.110 align:middle line:84% know that Mary passed and you don't know anything else. 01:06:59.110 --> 01:07:01.277 align:middle line:84% You might have imagined that it would work that way, 01:07:01.277 --> 01:07:02.350 align:middle line:90% but it doesn't. 01:07:02.350 --> 01:07:04.940 align:middle line:84% So if you ask me "How did people do on the test? and I say, 01:07:04.940 --> 01:07:07.785 align:middle line:84% "Well, Mary passed," there are lots 01:07:07.785 --> 01:07:09.910 align:middle line:84% of possible interpretations, but one is Mary passed 01:07:09.910 --> 01:07:12.250 align:middle line:90% and no one else did. 01:07:12.250 --> 01:07:14.380 align:middle line:84% So you do draw conclusions about everybody 01:07:14.380 --> 01:07:18.970 align:middle line:84% else partly from the fact that I failed to answer your question. 01:07:18.970 --> 01:07:21.760 align:middle line:84% That's one of the kinds of things pragmatists talk about. 01:07:21.760 --> 01:07:23.860 align:middle line:84% But it's possible that the particular tune 01:07:23.860 --> 01:07:26.800 align:middle line:84% that I sing as I say "Mary passed," 01:07:26.800 --> 01:07:31.030 align:middle line:84% invites you to contrastively focus Mary. 01:07:31.030 --> 01:07:33.190 align:middle line:84% That you're saying-- I'm saying to you, in a sense, 01:07:33.190 --> 01:07:35.830 align:middle line:84% Mary passed, and all the alternatives to Mary, 01:07:35.830 --> 01:07:38.080 align:middle line:84% notice that I haven't told you anything about them, 01:07:38.080 --> 01:07:39.520 align:middle line:90% draw your own conclusions. 01:07:39.520 --> 01:07:41.080 align:middle line:90% That's what I'm doing. 01:07:41.080 --> 01:07:43.810 align:middle line:84% Sometimes called a contrastive topic. 01:07:43.810 --> 01:07:46.240 align:middle line:84% And I think that's connected with your example as well, 01:07:46.240 --> 01:07:49.610 align:middle line:84% maybe, or something similar going on. 01:07:49.610 --> 01:07:51.053 align:middle line:90% No? 01:07:51.053 --> 01:07:52.970 align:middle line:84% As you might tell, there's a huge can of worms 01:07:52.970 --> 01:07:55.475 align:middle line:90% right here, this work on-- 01:07:55.475 --> 01:07:56.743 align:middle line:90% a really fascinating can. 01:07:56.743 --> 01:07:58.160 align:middle line:84% I shouldn't call it a can of worms 01:07:58.160 --> 01:07:59.360 align:middle line:84% because that makes it sound disgusting. 01:07:59.360 --> 01:08:01.250 align:middle line:84% It's fascinating, really, really interesting. 01:08:01.250 --> 01:08:02.625 align:middle line:84% There's all this interesting work 01:08:02.625 --> 01:08:05.720 align:middle line:84% on intonation, the kind of thing Faith is asking about, 01:08:05.720 --> 01:08:09.270 align:middle line:84% and the semantics of these kinds of expressions. 01:08:09.270 --> 01:08:11.373 align:middle line:84% So how do you connect the games you 01:08:11.373 --> 01:08:13.040 align:middle line:84% get to play with the pitch of your voice 01:08:13.040 --> 01:08:16.319 align:middle line:84% and your loudness with other sentences 01:08:16.319 --> 01:08:17.819 align:middle line:84% that I want you to think about that. 01:08:17.819 --> 01:08:20.689 align:middle line:84% I am not, in fact, saying but I'm communicating something 01:08:20.689 --> 01:08:22.729 align:middle line:90% to you about them. 01:08:22.729 --> 01:08:25.130 align:middle line:84% This is all fascinating, difficult stuff 01:08:25.130 --> 01:08:27.410 align:middle line:90% that people work on. 01:08:27.410 --> 01:08:27.910 align:middle line:90% Cool. 01:08:27.910 --> 01:08:32.458 align:middle line:90% 01:08:32.458 --> 01:08:34.250 align:middle line:84% All right, we're back from bats to cookies. 01:08:34.250 --> 01:08:35.810 align:middle line:90% Any questions about cookies? 01:08:35.810 --> 01:08:37.050 align:middle line:90% Bats? 01:08:37.050 --> 01:08:37.830 align:middle line:90% Only even? 01:08:37.830 --> 01:08:39.899 align:middle line:90% OK. 01:08:39.899 --> 01:08:43.410 align:middle line:90% Yes, association with focus. 01:08:43.410 --> 01:08:45.458 align:middle line:84% And then this is reminding ourselves 01:08:45.458 --> 01:08:48.000 align:middle line:84% that there are also sentences where you just focus something. 01:08:48.000 --> 01:08:52.817 align:middle line:84% "I will give Mary three cookies" used when alternatives to Mary 01:08:52.817 --> 01:08:55.109 align:middle line:84% are salient, like "Who will you give three cookies to?" 01:08:55.109 --> 01:08:57.180 align:middle line:90% I'll give Mary three cookies. 01:08:57.180 --> 01:08:59.183 align:middle line:84% Or some of you just said, "Are you 01:08:59.183 --> 01:09:00.600 align:middle line:84% going to give John three cookies?" 01:09:00.600 --> 01:09:02.700 align:middle line:84% And I say, "No, I will give Mary three cookies." 01:09:02.700 --> 01:09:04.977 align:middle line:84% Or when you're making lists, like "I'll give Susan 01:09:04.977 --> 01:09:06.810 align:middle line:84% three cookies, I'll give John three cookies, 01:09:06.810 --> 01:09:08.227 align:middle line:84% and I'll give Mary three cookies." 01:09:08.227 --> 01:09:12.180 align:middle line:84% So I'm inviting you to consider various alternatives to Mary. 01:09:12.180 --> 01:09:17.750 align:middle line:84% That's the point of putting this extra focus on these words. 01:09:17.750 --> 01:09:19.910 align:middle line:84% And then languages vary a lot, actually, 01:09:19.910 --> 01:09:22.082 align:middle line:84% with respect to how they realize focus. 01:09:22.082 --> 01:09:23.540 align:middle line:84% So there are languages like English 01:09:23.540 --> 01:09:28.010 align:middle line:84% that do it with gymnastics of the vocal tract. 01:09:28.010 --> 01:09:30.920 align:middle line:84% Your voice gets louder and softer, and higher and lower 01:09:30.920 --> 01:09:31.520 align:middle line:90% in pitch. 01:09:31.520 --> 01:09:35.120 align:middle line:84% There are other languages that tend to move the words around. 01:09:35.120 --> 01:09:37.700 align:middle line:84% So apparently, at least in some dialects of Spanish, 01:09:37.700 --> 01:09:39.470 align:middle line:84% a standard response to "Who bought 01:09:39.470 --> 01:09:40.819 align:middle line:90% the newspaper yesterday?" 01:09:40.819 --> 01:09:43.609 align:middle line:84% is literally "Yesterday bought the newspaper Juan." 01:09:43.609 --> 01:09:47.729 align:middle line:84% That is, you put the answer to the wh-question at the end. 01:09:47.729 --> 01:09:50.930 align:middle line:84% That's a normal way to answer these kinds of questions, 01:09:50.930 --> 01:09:55.430 align:middle line:84% although that's not the default Spanish word order. 01:09:55.430 --> 01:09:58.160 align:middle line:84% Or similarly, in Tagalog, if I want to say, 01:09:58.160 --> 01:10:01.880 align:middle line:84% "I'll only eat balut," you can't literally say that. 01:10:01.880 --> 01:10:05.750 align:middle line:84% So the last sentence on the slide is the attempt 01:10:05.750 --> 01:10:07.910 align:middle line:84% to translate word-for-word I will only eat balut, 01:10:07.910 --> 01:10:11.248 align:middle line:90% [NON-ENGLISH] 01:10:11.248 --> 01:10:13.040 align:middle line:84% I was doing some work with Tagalog speakers 01:10:13.040 --> 01:10:14.498 align:middle line:84% not too long ago trying to get them 01:10:14.498 --> 01:10:16.540 align:middle line:84% to accept this last sentence, and they were like, 01:10:16.540 --> 01:10:17.900 align:middle line:90% no, no, you can't say that. 01:10:17.900 --> 01:10:19.250 align:middle line:84% You have to say the first sentence, which 01:10:19.250 --> 01:10:20.708 align:middle line:84% is something literally like, "Balut 01:10:20.708 --> 01:10:22.260 align:middle line:90% is the only thing I will eat." 01:10:22.260 --> 01:10:26.450 align:middle line:84% So Tagalog speakers can't just say balut really loud 01:10:26.450 --> 01:10:28.700 align:middle line:84% and have that be the thing that associates with focus. 01:10:28.700 --> 01:10:30.740 align:middle line:84% They have to put it somewhere else 01:10:30.740 --> 01:10:32.900 align:middle line:84% at the beginning of the sentence in order 01:10:32.900 --> 01:10:33.960 align:middle line:90% to get it to mean things. 01:10:33.960 --> 01:10:36.620 align:middle line:84% Balut, if you remember, is this Filipino delicacy 01:10:36.620 --> 01:10:39.020 align:middle line:84% that involves a fertilized duck's egg, which is 01:10:39.020 --> 01:10:40.670 align:middle line:90% hard boiled before it hatches. 01:10:40.670 --> 01:10:42.467 align:middle line:84% Good source of protein, I'm told, 01:10:42.467 --> 01:10:43.550 align:middle line:90% though I've never had one. 01:10:43.550 --> 01:10:46.042 align:middle line:90% 01:10:46.042 --> 01:10:47.500 align:middle line:84% And then there are also languages-- 01:10:47.500 --> 01:10:51.850 align:middle line:84% this is Chickasaw, which is a Muskogean language spoken 01:10:51.850 --> 01:10:54.630 align:middle line:90% in Oklahoma, I think. 01:10:54.630 --> 01:10:57.520 align:middle line:84% It used to be spoken around Alabama, 01:10:57.520 --> 01:11:02.380 align:middle line:84% but I think they were moved, in which 01:11:02.380 --> 01:11:04.540 align:middle line:84% there are morphemes that you put on nouns 01:11:04.540 --> 01:11:06.220 align:middle line:90% that say this is the focus. 01:11:06.220 --> 01:11:09.530 align:middle line:90% 01:11:09.530 --> 01:11:11.960 align:middle line:90% We are just about out of time. 01:11:11.960 --> 01:11:14.133 align:middle line:84% There's one other topic on here, and I 01:11:14.133 --> 01:11:16.050 align:middle line:84% think we don't have time for the other topics. 01:11:16.050 --> 01:11:17.100 align:middle line:90% So I think we should stop here. 01:11:17.100 --> 01:11:18.620 align:middle line:84% Are there any questions about any of this, 01:11:18.620 --> 01:11:19.770 align:middle line:90% association with focus? 01:11:19.770 --> 01:11:23.180 align:middle line:84% So please ask me any questions you have left about semantics 01:11:23.180 --> 01:11:24.980 align:middle line:84% because I refuse to teach you anything 01:11:24.980 --> 01:11:26.430 align:middle line:90% about semantics from now on. 01:11:26.430 --> 01:11:28.580 align:middle line:84% We will never talk about it again. 01:11:28.580 --> 01:11:29.570 align:middle line:90% That's probably a lie. 01:11:29.570 --> 01:11:31.910 align:middle line:84% I mean, I was telling you about syntax, and phonology, 01:11:31.910 --> 01:11:33.550 align:middle line:84% and morphology today, and those are 01:11:33.550 --> 01:11:35.300 align:middle line:84% things we thought we had left behind, too. 01:11:35.300 --> 01:11:37.085 align:middle line:90% It all kind of blends together. 01:11:37.085 --> 01:11:40.550 align:middle line:84% Any last-minute questions about semantics? 01:11:40.550 --> 01:11:42.350 align:middle line:84% All right, go out and enjoy the day, 01:11:42.350 --> 01:11:46.450 align:middle line:84% and we'll see you guys on Tuesday. 01:11:46.450 --> 01:11:51.000 align:middle line:90%