1Week 3a. Categories, features,natural classes, and morphology.CAS LX 522Syntax IPreviously, in LX522…n So, here’s where we were.n We’re trying to characterize our knowledge ofsyntax, using English speaker’s knowledge ofEnglish as a window to the kinds of things weneed to describe language.n Words seem to come in categories (N, V, A, P, C,I, PRN, D, …). English treats these differently, soto describe English, our theory needs to treatthem differently.Previously, in LX522…n We’ve collected a number of features thatseem to make a difference in variousways.n We have the category, e.g., [+N], [+V].n But we seem to have subcategories too,e.g., [+Pl], [+Common], [+Count].n These features matter to how we cancombine words. Language cares, and so,therefore, do we.The surfacen Before going on, let’s take a detour,because if we don’t, things are just goingto get confusing.n So far, we have looked at a word, and triedto determine what its relevant featuresare.n Books [+N, +Count, +Common +Pl].n Written [+V, +Participle, +Perfect]Where’s the feature?n This is useful in that we get a hint as to whatfeatures are required. But consider:n Bill ate lunch.n Bill will eat lunch.n Bill did not eat lunch.n Bill does not eat lunch.n You do not eat lunch.n We do not eat lunch.n What are the features of ate? eat? did? does? do?Why is this so confusing?n So, do seems to be:n [+Pres], andn not [+3, +Sg], andn only shows up in the negative.n Ouch.n That’s rather inelegant.n Here’s the problem: I sat on the bank. I saw thecandidate with the binoculars. Visiting relatives canbe tedious.n See?2We’re going the wrong wayn There are two different intents underlyingVisiting relatives can be tedious (so I do it aslittle as possible), and Visiting relatives can betedious (so I avoid them as often as possible).n They happen to sound the same, but theyhave a different underlying structure.n In general, what’s unique is theunderlying intent/structure, not thepronounced form.Generative grammarn The syntactic system we are going to build is agenerative grammar.n It builds up an underlying structure, which isthen pronounced.n The two versions of Visiting relatives can betedious are different sentences.n …But wasn’t our goal to explain how peoplecould tell if sentences they hear are part of theirlanguage or not?Judging sentencesn The view of sentence judgment we’lladopt here is basically one of askingoneself: Could I say that sentence?n When listening to somebody, you ofcourse need to decode what that personmeant, but it is a process of recovering theunderlying form of their utterance.A very, very little bit of Frenchn If you’ve tried to learn any French at all, you’vecome across this phenomenon:n de ‘of’ le ‘the (masculine)’n à ‘at’ la ‘the (feminine)’n à la biblioteque ‘to the library (fem)’n *à le cinéma ‘to the movies (masc)’n au cinema ‘to the movies (masc)’n de la mayonnaise ‘of mayonnaise (fem)’n de le lait of milk (masc)n du lait ‘of milk’ (masc)A very, very little bit of Frenchn This is usually taught as:n au = à + len du = de + len If your underlying intent is à ‘at’ + le ‘the’,you pronounce it like au.n So is au a preposition or an article?What does this have to do witheating lunch?n And now we can return to the point:n Bill ate lunch.n Bill eats lunch.n Bill does not eat lunch.n Bill will (not) eat lunch.n What generalizations can we come up with here?How are the features organized in these simplesentences?n Why did I juxtapose à+le=au from French with Bill atelunch? Where is tense? Where is the verb? What is ate?What does do mean?3Falling into placen If we suppose that these sentences all have thesame form…Subject Tense/Agreement (Not) Verb Object…things start to look a lot more regular,describable. This is the structure of a sentence.n That Tense+Verb comes out as “Tensed Verb” is amatter of pronunciation. If you separate Tense fromthe Verb with not, they no longer can combine. Inorder to pronounce Tense, you insert do.Falling into placen Moving one step closer to syntacticstructure:[NP Subject] I (not) V [NP Object]n So [±Past, ±1, ±2, ±Pl] are features of I.n [±1, ±2, ±Pl, ±Common, ±Count, …] arefeatures of N.n [+V] is a feature of V.Those pesky participlesn Bill will have been eating lunchn [NP Bill] [I will] [V have] [V been] [V eating] [NPlunch]n will [+Fut]n have [+Vaux]n been [+Vaux, +Participle +Perf]n eating [+V, +Participle, -Perf]Crosscategorial featuresn Consider what un can attach to.n untie, unfold, unwrap, unpackn unhappy, unfriendly, undeadn *uncity, *uncola, *unconventionn *unupon, *unalongside, *unatn Basically, it applies to reversible verbs andadjectives, but not to nouns orprepositions. How can we state that?Crosscategorial featuresn Suppose that nouns and verbs are the most basiccategories. A noun is a noun and not a verb, andverb is a verb and not a noun.n Noun: [+N, -V].n Verb: [-N, +V].n A possible conceptual reason to separate nounsand verbs is that verbs are basically predicates—they attribute some property to the noun. Nounsare basically arguments, to be assigned propertiesby verbs.Crosscategorial featuresn Looked at this way, adjectives are kind of“verby” in that they are also attributingproperties.n It’s hard to make that really precise, but wehave a more concrete syntactic similaritybetween verbs and adjectives too: both cantake un-, while nouns and prepositions cannot.4Supercategoriesn Chomsky (1970) proposed that we explainthis by supposing that [±N] and [±V] arethe two basic features that determine thefour lexical categories (N, V, A, P).n N: [+N, -V] V: [-N, +V]n P: [-N, -V] A: [+N, +V]n Given that, what does un attach to?Supercategoriesn So, un attaches to a [+V] category. Itdoesn’t care about [±N]. [+V] defines anatural class that language refers to.n Why is A [+V, +N] and P [-V, -N]?n Suppose we had a morpheme that attachesjust to V and P, how could we state that?n Do V and P form a natural class too?Russian Casen Other languages can give us evidence of naturalclasses as well. E.g., Russian nouns (all nouns)are marked for Case (like English pronouns are:me vs. I), but when they are modified by anadjective, the adjective is also marked for case.n What gets marked for Case in Russian?‘The beautiful girl put the black cat in the empty box’boxemptyincatblackputgirlbeautifulkorobkupustuyuvkoshkuchornuyuvsunuladyevushkaKrasivayaFunctional and
View Full Document
![Nom from finite T no case from nonfinite T (T [inf], M to)](https://media.gradebuddy.com/documents/386389/7a3e15e0-7507-4825-a32d-f3e6fb67d2eb.png)
Unlocking...