On 9/13/07, <b class="gmail_sendername"><a href="mailto:maxwell@umiacs.umd.edu">maxwell@umiacs.umd.edu</a></b> <<a href="mailto:maxwell@umiacs.umd.edu">maxwell@umiacs.umd.edu</a>> wrote:<div><span class="gmail_quote">
</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Rob Freeman wrote:<br><br>> I think the syntax of "supported" in each context will be different.
<br><br>What is your definition of "syntax"?</blockquote><div><br>John doesn't like one sentence summarizations. But I do, so one might be:<br><br>Generalizations (ad-hoc) about the way words can combine.<br>
</div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">> Failure to treat the syntax of "supported" on an ad-hoc basis in this way
<br>> means you have no way of capturing the information that, in a grammar of<br>> English, "supported" = "accompanied" but also "supported" !=<br>> "accompanied".<br><br>
The question is whether it's appropriate to capture that meaning<br>distinction in the syntax, rather than in the semantics. It seems to me<br>that there's a class of facts that can be captured quite well by<br>distinguishing syntax from semantics, but which is obscured when you try
<br>to collapse the two.</blockquote><div><br>You gave some examples of this earlier, didn't you? At least, you gave some examples to demonstrate syntax is important. I wasn't bothered then because I also believe syntax is important. Now I'm trying to argue syntax is not only important, but that it can code meaning.
<br><br>You'd better give the examples again.<br></div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">> That means you will be unable to capture detailed syntactic restrictions
<br>> which prevent you from saying "slightly odd" things like "Tom accompanied<br>> his tomato plant to the garden (where he planted it.)"<br><br>I would put it differently; I would say that collapsing syntax and
<br>semantics makes it difficult (maybe impossible :-)) to capture the fact<br>that the above sentence is perfectly comprehensible, if (as you say) odd;<br>whereas the following sentence is only comprehensible to English speakers
<br>with great difficulty, although the intended meaning is quite normal:<br> garden-the-to accompanied tomato-plant-his Tom</blockquote><div><br>I think you are trying to demonstrate that this "sentence" has meaning, but is incomprehensible because it does not obey English syntax.
<br><br>By which you probably hope demonstrates the independence of syntax and semantics.<br><br>Which is a slightly moot point. I'm not trying to demonstrate meaning cannot exist without syntax. It can. I'm trying to show syntax can code meaning.
<br><br>You have less (recognizable) syntax in your example, so the we lack that code. That does not mean a recognizable syntax would not code meaning, only that we are forced to use what we do have (lexicon, and a bit of pig-Warlpiri?) to guess the meaning a missing syntax might have coded, in this case.
<br><br>If I take your example one stage further and show you a _picture_ of Tom carrying the plant out into the garden, does that then demonstrate _language_ does not code meaning (because you don't need language to convey meaning)?
<br></div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">> John would claim such restrictions are purely semantic, but in point<br>> of fact you can capture them with an ad-hoc search for syntactic
<br>> regularities along the lines I recommend.<br><br>I'm sure you can find such regularities; the question is whether treating<br>them as syntactic doesn't obscure the generalizations that I would<br>consider truly syntactic.
</blockquote><div><br>By "truly syntactic" you mean syntactic generalizations which don't code any corresponding meaning?<br><br>Let's look at them. Maybe such things exist.<br><br>It's possible. Remember my argument is slightly different from the traditional. This kind of debate has usually been framed in terms of whether semantics governs syntax. If that were true there would be no syntactic distinctions which did not have a semantic basis. But I'm not arguing that. I'm not saying semantics governs syntax (John is saying that, if anyone.) I'm saying syntax, which you can define independently in terms of regularities in texts, can be used to code semantics. (Though you don't need it. You don't even need language. You might equally paint a picture and get the meaning across that way.)
<br><br></div>-Rob<br></div>