The past few years have witnessed a shift in reasoning in how traditional grammar should be conceptualized. This shift, I believe, has done well to naturally aid students in achieving a higher and more comprehensive level of language. The aim of this companion handbook is to provide an elementary introduction to recent developments in syntactic theory--particularly working within the framework of Chomsky's 1995 Minimalist Program. More specifically, the handbook focuses on a theory called Feature Theory, as it has to do with basic levels of grammar. Although Feature Theory is an integral part of Chomsky's overall theory stated within the Minimalist Program, there is nothing inherent in the theory itself which should prevent it from being presented alongside, say, other textbooks on the topic of grammar which in fact may correlate to other syntactic theories. In other words, the principles behind Feature Theory as presented herein are understood to be based upon universal characteristics of all languages--characteristics which transcend all common discussion of grammar. For example, recent work on Features has refocused attention on traditional distinctions placed on Form Class Words vs. Structure Class Words (and more specifically, Lexical vs. Functional Categories). The core of this text attempts to provide students with a good working knowledge of such features as they have to do with the more formal aspects of functional grammar, and to allow students to utilize this working knowledge to build "syntactic trees" (diagramming) one feature at a time. Ultimately, the hands-on work will provide students with an inside peek at the multi-layered fine structure of grammar--starting with the more primitive, basic foundations of what makes a simple sentence to the unraveling of those finer grained features which form the makings of complex functional grammar.
0.
Grammar (The asterisk "*" throughout indicates an ungrammatical sentence). Recall, the derivational process sketched out above has taken the main Verb stem {teach} and changed it into a Noun {teacher}. Surely, this change from Verb to Noun has an immediate effect on how we are able to construe the word in a given sentence. In short, (postponing further discussion to later sections), the syntax involved here would be the following: (0) [Subject] (John) + [Finite Verb] (like-s) + (optional) Infinitive verb complements:
The syntax
doesn't allow the option of an infinitive verb marker {to-}to attach
to nouns*[{to} + [Noun] ]. It is precisely this infringement that
makes the sentence illicit.
Although all
languages have words, and the word is typically regarded as the
sacred unit of meaning that drives all of language, there is a considerable
amount of linguistic material that cannot be neatly packaged into
a "layman's" notion of word. For instance, it is argued
that one doesn't learn words as isolated word islands. Rather, it
seems that one learns words in the following two-prong manner: (i)
as words relate to meaning (lexico-semantics)--based on
a one-to-one relationship of sound-to-meaning, and (ii) as words
relate to word classes (lexico-syntactic)--based upon where
the word sits within a sentence. So overall, all three linguistic
branches of study are ultimately involved with the learning of the
basic word: Phonology (sound), Morphology
(meaning), and Syntax (class). (See §0.3 for
the role of syntax in word learning).
0.1 The "Sally Experiment": An Introduction of Lexical vs. Functional Grammar One very nice way to illustrate the essential difference found between Lexical and Functional grammar is to call upon an experiment referred to here as the "Sally Experiment" (Galasso 1998, class lectures: Univ. of Essex). The experiment offers us a classic case of how ESL students tend to realize units of grammar (ESL=English as a Second Language).The token 'Sally' sentence below illustrates in a very natural way the classic distinction made between what is Lexical vs. Functional--a distinction typically referred to as Substantive vs. Non-substantive units of language. The heart of the experiment relies on the distribution of the /s/ in the sentences below: Sally wears strange socks. (1)
It should be made obvious in the token sentence pair (one of many presented in the experiment) that the phonological unit (or phoneme) /s/ is what is being examined here. However, when one takes a closer look, there emerges an interesting asymmetry in "what gets left out where" in specific ESL contexts (ex. 1b). It should be said that on the phonological level, all /s/'s throughout are relatively the same--that is, they are similarly pronounced (notwithstanding some r-voicing assimilation that changes the /s/ to /z/ in the word wear-s). So, an account for the apparent asymmetric distributions of /s/ cannot be made on the grounds of phonology. In the case above, it appears that although ESL students may pronounce correctly and produce 100% mastery of the underlined phoneme /s/, they tend to optionally omit the italic /s/. This forces early-on in our discussion of grammar a further distinction between (i) Phonology, on the one hand, and (ii) Morphology, on the other. For example, if all underlined /s/'s are produced 100% of the time, surely, as expressed above, there is no phonological deficit. The optional omission of final /s/'s here must be attributed to a deficit in morphological. Hence, the two aspects of grammar are addressed simultaneously--Phonology vs. Morphology and Lexical vs. Functional: The lexical /s/ being the one underlined and the functional /s/ being the one in italics. These two very distinct aspects of language (and language processing in the brain) introduces us to a very important and seemingly transcendent dichotomy in language--viz., Lexical vs. Functional Categorical Grammar (as illustrated below): (2)
What we see in the sentence
experiment above, and expressed in the diagram in (2), is that the
lexical /s/ is never dropped. This phenomena can be accounted for
by the fact that the lexical categories--here being a lexical item
/s/--are composed of crucial substantive (lexical) information and
must be preserved in order to effectively communicate the whole
lexical/word meaning. For example, the initial /s/ dropped in Sally
would give us ø-ally /æli/ (in IPA), which would completely
distort the intended meaning. The same problem would arise with
ø-cks /aks/. In these cases, the /s/ is said to be lexical
because it contributes to the overall word meaning: without the
full lexical meaning to which the /s/ contributes, the meaning is
changed. On the other hand, and in contrast to the lexical /s/,
if the functional /s/ is omitted, there occurs no meaning loss.
Functional elements of a given sentence can therefore be defined
as being "non-crucial" for the actual transmission of
communication. Whether or not we say "wear" or
"wear-s" tells nothing of the actual meaning
of the word--viz., the /s/ in "wear-s" must be
present only to carry out an abstract relationship of functional
grammatical between (i) Sally [Pronoun: 3Person/Singular]
and (ii) wear-s [Main Verb: 3Person/Singular/Present].
0.2 Structure vs. Form Class: "How do you do?" In additional to the Lexical vs. Functional category distinction at the morphological-inflection level, the same distinction holds at the word level: the distinction is labeled (i) Form Class word vs. (ii) Structure Class word. One way of observing this lexical vs. functional distinction at the word-level is by considering the token interrogative sentence "How do you do?", where the obvious double usage of the word "do" should stand out. In fact, in some of my years of teaching abroad, I have even had the question posed to me in the following manner--"What does the second "do" mean and why do we have to repeat it so"? The question stands to an extent only insofar as it depends on the misunderstanding that--if the two words have identical meaning, then how come the repetitive nature of the phrase. As we shall see later on in this text, the two "do's" are indeed not one in the same (notwithstanding the perceived identical pronunciations). Herein lies the confusion: The first "do" is actually functional, containing no meaning whatsoever and only serves some abstract functional purpose--here, it specifically serves to form the grammar of a question (interrogative) sentence (See (ex. 110) and following regarding the Auxiliary Verb and it functional role in grammar). It is only the second "do" which is lexical and thus contains very general generic verb meaning (as in the verbs go or feel in the greetings "How's it going?, How do you feel?, etc.). One simple way to uncover this distinction between lexical "do" and functional "do" is to evoke the substitution test--a beloved test of linguists which often helps to get a better handle on the nature and distribution of a particular class or category of words. Consider the substitution test below in (3) where we can see the selective distribution between (i) the first Functional Auxiliary-Verb "do" (Verb1) and (ii) the second Lexical Main-Verb "do" (Verb2): (3) Table: The Substitution Test
(* marks ungrammatical sentence, @ acceptable in reduced speech, ex. j) Surely, "* How speak you do?" (3g) is an improper, ungrammatical interrogative sentence. This distinction goes to the heart of the issue as discussed above. By misplacing the verbs into the opposing slots, we shatter the syntax and thus the overall meaning of the sentence. More specifically, the lexical "do" (which is always positioned in the Verb-2 slot with regards to interrogative sentences) is the main verb and carries the substantive meaning of the verb, whereas the functional "do" (Verb-1)--sometimes referred to as the "Dummy-'do' insert"--is merely an Auxiliary verb (void of any verbal meaning) and is inserted between the Wh-Question and the Subject in the capacity of an abstract interrogative marker. (See Form Class vs. Structure Class). This is precisely why sentences 3f-i are ungrammatical--namely, where we ought to have a substantive main verb carrying out its full verbal meaning in the appropriate slot, we have instead a 'Dummy-do' auxiliary verb void of any potential meaning. Returning then to the original question which spawned the above substitution test, we now see that indeed the two seemingly identical "do's" are not alike--whereas they may be alike on a phonological level /du: /, they are two very different items at a morpho-syntactic/grammatical level. (Note that in fast pronunciation, the first Aux "do" gets reduced to /hau-y-du/ (IPA) (=How ø you do?).
The natural first steps in attempting to systematically categorize language (in general) would be to (i) establish a natural class of word types (e.g., Nouns and Verbs) and (ii) define such word types as containing common word-level and distributional features. Much of this information regarding "word types" is already in the minds of speakers--it's part of our endowed linguistic knowledge given to us (free) at birth. However, one major contention surrounding this assumed innate source of language knowledge is the hypothesis that the brain, therefore, must house, in isolation, some special (built-in/native) autonomous module for language, disconnected and disassociated from all other modules in the brain which might lead back to general cognitive skills, etc. This school of thought is known as Special Nativism (as opposed to General Nativism which assumes a Piaget-style language learning process tethered to more general cognitive development). (Return to my language schema diagram in (2)). Let's take a quick peek into such "built-in" knowledge by considering how native speakers can manipulate novel words in the following sentences below. Consider how a novel word "Sib" (Brown, 1957) (a newly created word not part of our English input) takes on appropriate syntactic categorical status: (4) Table: Word Category & Rule
Of course, on a Semantic level (or word meaning-level) you don't know what the word sibbing actually means (e.g., This 'sibbing'/'F'-ing car!!!) (we can leave telepathy to work here and so any number of suggestions is open to the floor). However, due to sib's syntactic and grammatical distributional properties, one can (i) infer some amount of meaning while (ii) attributing a categorical status simply by tracking the word through the overall sequence of the sentence. In other words, on a basic level of discussion, one could say that we arrive at word meaning via its placement (where it sits in contrast to other words in a sentence). This placement is syntax, and this approach to lexical/word learning is known as Syntactic Bootstrapping. It is not too far fetched to assume that the word class to which we have attributed "sib" is specified in that word's lexical entry: [+/-N, +/-Adj]. A grammatical category is thus a class of words which have a common set of grammatical features. The traditional "category" basis for defining words as "parts-of-speech"--namely, Verb/Adverb, Noun/Adjective, Preposition--has been fundamental throughout linguistics. Verbs and Nouns are the two highest profile categories which enter into a wide range of grammatical relations: viz., most Nouns enter into a grammatical relationship showing e.g., (i) Definiteness distinction (A book vs. The book), (ii) Number distinction (Singular vs. Plural) (A/The Car, *A/The Car-s), etc., while Verbs enter into a full range of forms termed Inflection (or Tense/Agreement): (5) Table: Verb Forms: Inflection & Grammar
Let us then take as a basic starting point the following criteria for determining a Noun from a verb: the one essential defining and distinguishing factor between, say noun vs. verb is that nouns can take-on plural {-s} (and not verbs), while verbs can take-on past tense {-ed} (and not nouns). This is of course oversimplified, but for the time being it should serve us well. In addition to the full range of forms Nouns and Verbs receive, at the isolated word level, there are other differences which appear at higher syntactic levels: e.g., (i) Determiners introduce Nouns, and (ii) Auxiliary/Modals introduce main Verbs. The aspects of functional categories--in this case Determiners and Auxiliary/Modals--specifically addresses this notion of a Lexical vs. Functional relationship. All Lexical Nouns and Verbs which convey semantic/substantive meaning are "helped" in maintaining their abstract (functional) grammaticality by their own counterpart functional co-host: Nouns by Determiners, and Verbs by Auxiliary/modals. (Note that auxiliary verbs "Do-Be-Have", which must work in conjunction with main lexical verbs, are often called "helping verbs"). Recall, that if lexical categories contain only mere semantic material, and little if any grammatical material, then in order for them to enter into a true grammatical arrangement (syntax), they need to take-on abstract grammatical features derived from their functional counterparts. This duality between Lexical & Functional categories goes to the heart of how abstract grammar emerges (recall the schema in (2)). While the full range of Functional Features will be spelled out in the ensuing sections, let's briefly introduce the notion here. Let it suffice for now to say that it's the Functional Determiner The that renders the Noun book specific--as opposed to the generic A book. This distinction being played out here relates to a specific functional feature that has to do with Def(initeness): namely, a [+Def] Feature carried by the Determiner and thus affecting the counterpart Noun: e.g., The book vs. A book. (See (48) below for Def-features). Here, it is the binary realization of either a [+/- Def] Feature which can be attributed to the distributions of Definite [+Def] vs. Indefinite [-Def] Determiners (The vs. A respectively). Consider in (6) below the syntax between lexical nouns & verbs in how they enter into functional relationships between determiners & auxiliary/modals (respectively). (6) Table: The Syntactic Range of Nouns and Verbs
A category based model of language classifies words according to parts-of-speech. For example, note that a word such as joke would take a plural {s} (forming the conjunct joke-s) => word category [+Noun/-Verb] (because only nouns can incorporate the plural {s} inflection), but not the word e.g., jokingly (*jokingly-s) (asterisks * marking ungrammaticality). Whereas the former word joke is categorized as [+Noun], the latter word jokingly is categorized as [-Noun, +Adverb]. Such basic categorization is well and good at one level of investigation; however, as we shall see below, such over-simplified labeling based on pure categorization becomes insufficient and problematic when faced with more subtle distributional properties that accompany words. Although this basic model of representing words (via parts-of-speech) intuitively assumed that there had to be something in the internal make-up of the lexical items (per se) that either allowed or disallowed certain types of inflections, no real attempt was made to account for differing behaviors and distributions of particular words of the same class. In other words, while a category-based model provided a broad description of the bundle of words which made-up a category (N, V, Adj, Adv), it provided no outlet for describing crucial differences found between words within the same category: e.g., traditional notions of Count vs. Non-Count (Mass) Nouns (Two book-s /teacher-s/ home-s/ *furniture-s). Here, the word furniture surely is grouped and classified as [+Noun] as are the rest; nevertheless, it doesn't take the plural {s} inflection. There seems to be no way to handle the distinction of the two types of nouns short of separating them into a separate dual-category (e.g., [+/-Mass] vs. [+/-Count]). While this "dual-category approach" would certainly capture this refined difference between the two nouns, it would do so as the expense of abandoning their larger and more important categorical "[+Noun]" class similarity. What seems to be needed in this case are finer grained models which (i) divert lexical analyses somewhat away from the actual word-class item itself (but not at the expense of the class), and rather (ii) examine the possible arrangement of the precise lexical internal features (i.e., sub-categorical features) which factor in such differences between words of the same class. On the heels of such analyses, a methodological paradigm shift has occurred which defines "Word-bundles" as "Feature-bundles"--using binary notation to express the full range of properties (and feature-matrices) inherent in a given word. This new approach couples (i) the traditional category-based model which sets out to define general parts-of-speech with (ii) the more subtle feature descriptive model. (For further reading on Features, see the paper Non-count Noun Determiners: Where's the feature?" posted on my web site). (NB. Having said this, it is also important to note that this "Reductionist" view--that traditional syntactic categories can be broken down into smaller parts--shouldn't imply that the "labeled" category is now made redundant. The traditional category, as we have always known it, remains and is not replaced by say a bundle of features. It is perhaps better put by saying that the two go hand and hand. Both feature classifications--(i) Category-Features which seek to maintain some coherency with the traditional "parts-of speech", as well as their more subtle counterpart (ii) Subcategory-Features--are typically represented in enclosed brackets [ ] and may be expressed in binary notation [+/-F] (with (F) indicating the specified feature). For example, a Noun category may be defined by a binary expression denoting intrinsic category features on the one hand--such as [+N, -V], as well as subcategory-features [+Count] on the other. The above notations state that the lexical category Noun has an intrinsic plus N-feature and minus Verb-feature, in addition to having fine-grained functional sub-category plus count-features (number). This now nicely captures the distributional differences found among the words *furniture-s vs. book-s as witnessed above--with the former being notated as [+N/-Count] and the latter [+N/+Count]. (It is worth noting that this functional distinction between Count and Mass Nouns are not fully appreciated by children who may be just entering their Functional-Stage of language development (approx. 2-3 years of age): it's been well documented over the years that children both (i) combine indefinite determiners with mass nouns (e.g., a dirt, a sand) as well as (ii) over-generate plural {s} on mass nouns (e.g. I have two furniture-s, three sugar-s, etc. and that some children continue to make such errors of over-generalization up until eight years of age. (see Brown and Bellugi, 1964, Slobin 1966 for an overview). At the basic level, the four traditional parts-of-speech categories can be represented by the following binary notations: (7) Sub-categorical Features: Binary Notation
In addition to these inherent categorical features, their inherent Lexical vs. Functional status might also be notated in a similar way (as shown in (15).
Lexical Words have specific and idiosyncratic meaning. These words are content-based and can either be readily conceptualized in the mind of the speaker (i.e., semantically-based as with Nouns & Verbs) or can come to be manipulate upon in logical terms (i.e., logical no, and , and if/or) and/or potentially take-on opposite meanings (e.g. Adjective cold>hot, Adverb slow>fast, Verbs sleep>wake, etc.) The list of lexical categories is given in the table below:
These categories are expressions of lexical items--"Lexical" here simply means "Word" (as would be found in a dictionary with an assigned substantive-meaning definition). All lexical words share a common property of being content driven--that is, the word is anchored by some substantive meaning. In this sense, the speaker should be able to conceptualize the properties of a given lexical word--whether it be Noun (adjective) or Verb (adverb) (and to some extent Prepositions contain meaning albeit via a relatively positional relationship). (But see Preposition as Functional Category in §2.4 (161) below). In other words, the speaker should be able to construct some form of a mental image of say, folder, mailbox, ball , etc. (for nouns), or a mental action of say, dance, run, eat, etc. (for verbs). In a basic sense, we could assign some sort of meaningful iconic representation to the sound-meaning associations in the following nouns in (9). (Although verbs are to a large degree less salient in concrete terms (tangibles, etc.), they nevertheless contain meaningful conceptual information that is related to states or actions).
Notwithstanding the abstract and less salient nature of Main Verbs
as expressed above--viz., the traditional notion that Nouns represent
more readily accessible conceptual notions of person/place/thing--out
of all the lexical categories, the Preposition Class is
perhaps the most abstract when attempting to form a stable mental
image. However, its substantive nature does make itself available
to us via location and manner. Clearly, words
such as between, below, in, with, under, etc. have some
conceptual value in relation to structural, locative meaning. (See
§2.4 regarding functional features of Prepositions. Another interesting aspect of lexical category words is that they are the first type of words to be spoken by children in their earliest stages of speech. This stage-1 is typically referred to as the lexical-stage of language development. (See relevant papers on language acquisition posted on my web site). Form Class Words Consider the examples in Table (10) below of how the word-stem forms of our four main lexical categories change via Inflectional and Derivational Morphology: (10) Table: Form Class Lexical Words: Inflection & Derivation
For further exercise, analyze the English derivation morphology in the following words in Table (11) below. Try to identify the root-stem lexical word, along with the function of each derivational affix (creating the derived word): (11) Table: Exercise in English Derivational Morphology
(Examples taken from Radford et al. 1999, p. 177) What will be of interest to us in the following sections is the idea that Form-class Lexical words can take-on inflection whereas Structure-class Functional words cannot (see §0.3.2 and (157) on Modals). This distinction will later become a major theme in our overall grammar--in the sense that in order for "meaningful" lexical words to contain more "abstract" levels of grammar, they must allow their forms to change and be affected via Functional Inflection--this Inflectional Process, which is so much a part of what we understand (functional) grammar to be, will be more fully fleshed out in subsequent sections and chapters.
In contrast to lexical categories, which contain meaning, Functional Categories (or features) are a class of Words (or inflections) which have no substantive meaning, and are thus inserted into a sentence not to transmit tangible information, but rather to serve some abstract grammatical purpose--functional words or items (inflection) are usually utilized in some capacity to form a grammatical relationship with a counterpart lexical item. (For example, go to the DP (Determiner Phrase) example to see how a Functional Determiner "The" might work alongside a Lexical Noun "car": DP). In a sense, what we shall later see is that functional categories assist lexical categories in carrying out grammar. A list of the major Functional Categories is as follows:
Showing the contrast between the two categories, we could say that lexical categories have descriptive content, whereas functional categories have no descriptive content. For instance, whereas the noun "dog" (in the phrase The dog) can be easily conceptualized in the mind of a speaker, the determiner "The" cannot be so readily conceptualized. In fact, there is actually no meaningful content to the word the. For instance, just try to make an image in your mind of what the might look like (shape, size, color, action, etc.)--as you quickly discover, it is an empty search. This is because lexical words are stored in your mental-lexicon (a sort of memory file of words) in such as way as being labeled, associated and indexed to meaning--this is what is behind the notion of a 1-to-1 association (sound-to-meaning association or indexing (cf. Skinner). Though we shall more closely examine the roles of the major categories in later sections, let's briefly look at two main functional categories as listed above: D & Aux. The Determiner (D) class is a functional (or Structure Class) group of words which specifically work in conjunction with counterpart Nouns. Such prototypical members include: Articles (a, an, & the), Demonstratives (this, that, these, those), Possessive or Genitives (my, your, his, her, our, their, its), Indefinites (some, any, no, many), Cardinal Numbers (one, two , three...), and Ordinal Numbers (first, second, third...). What is important here to capture is that all the above words work in conjunction with Nouns--forming a functional/lexical relationship: e.g., [D A/The/This/My/One] + [N car]. Whereas all determiners share in this common relationship (D+N), specific determiner words also maintain their own special properties. The most common property of all is that (D)-words serve to introduce Nouns--plainly speaking, Determiners signal that a Noun is fast approaching within the phrase. In fact, except for very special grammatical conditions where it is possible to dispense with the Determiner--e.g., when the noun contains general, generic information as in the sentence ø Girls just want to have fun" showing no overt sign of a D-word (ø) to introduce the plural general class noun Girls--it is seldom possible to have a noun without a D-word introduction: e.g., *I like car/She is friend/Where are toys/This is not book/How do you like weather?/You need to study for test, etc. etc. (* marks ungrammaticality). And more to the point, it is never possible to have a D-word without a Noun: e.g., *I like the/She is my/Where are these?/This is not a/How do you like our ?/You need to study for your , etc. etc. (NB. There is a class of 'Determiner-like' counterparts that serve as pronominals and must stand alone. For example, consider the following possessive-determiner/(pre)pro-nominal paradigm: my/mine, your/yours, her/hers/, our/ours. For example, contrast the following--This is my book, This is mine ø vs. *This is my ø, *This is mine book. Those are our books, Those are ours ø vs. *Those are our ø. *Those are ours books. The D-words my/our here are considered to be prenominal in that they are required to come before nouns, while their counterparts mine/our are pronominal (non-determiners) and thus must be used in the manner of a (Pro)noun). If we wish to maintain the condition pronounced earlier that Determiners hold a structure dependency in that they must introduce Nouns, then such Pronominals as cited above can't be classified as Determiners--their distributional properties, as well as their syntactic behavior, hold the status of [+Pronoun]). Returning to their special properties, consider the determiner "my". Not only does this D-word (my) signal the presence of an ensuing Noun [My + N], it also expresses a special grammatical property of ownership or possession: as in the possessor + possession relation [D My] + [N car] (respectively)--"Hey, that car belongs to me--my car!" (as opposed to a generic, proto-class reference to "car" (e.g., "The/A/Some car"). So, it's easy to see the contrast between the two determiners my & the--although, on one level, they both introduce nouns, their grammatical properties are indeed very different. We will more closely examine the full range of specific features related to determines below. For the moment, the most important aspect of the D-word is that it works hand-in-hand with nouns. One trivial way to put it would be to say that the (oddball) determiner is to the (common) noun what "Dr. Jeckle" was to "Mr. Hyde"--in a sense, both are two sides of the same coin. Consider the diagram of a [D+N] below:
(* Using binary notation, the possessive D-word could be specified by a finer-grained [+Gen] feature that resides in the overall Determiner category rendering a (Gen)itive Possessive Determiner. See §2.1 (57) on Genitive Determiners). Structure Class Words
As introduced above, the entire range of lexical and functional categories can (more or less) be presented by a binary notation in which specific sub-categorical features and properties are indicated. The feature [+/-N] serves to cross-classify categories in a similar way--it implies that all [+N]-words (nouns) share a common property and thus form a super-category or class called Noun which differentiates the class from say [-N] words (such as Verbs, Prepositions). Likewise, we can use this same notation to account for the intricate Functional-to-Lexical inter-relations as noted above. It was noted that each Functional category closely works alongside a corresponding Lexical category: e.g., Determiner + Noun, and Aux/Modal +Verb. Consider the notation of the functional categories below where [+F] shows (Functional category). (15) Table: Word Category & Features
Recall that determiners may indeed precede a Verb--hence turning it into a Noun (=Gerund) (as shown in (16) below): (16) Table: Gerund Constructions
Gerund's particular use of {+ing} forms may create Noun counterparts to DP-Subjects/Objects, as well as modifying Adjectives as in e.g., My sleeping/white cat is fine. As stated above, the lexical/substantive categories--which provide meaning--have a functional categorical counterpart. The diagrams below help to illustrate this specific inter-relationship between the categories of three fundamental phrases: (DP, MVP, PP): (17) DP, MVP, and PP Phrase Diagrams a). Determiner + Nouns => Determiner Phrase (DP) 
b). Auxiliary + Verb => Auxiliary, or Main Verb Phrase (MVP) 
c). Preposition + [DP] => Preposition Phrase (PP)
d). Determiner + Pronoun => Determiner Phrase (DP)
[D= ø, N (Pro-Noun) = I]
As was mentioned earlier, the preposition (PP= Prepositional Phrase) seems to be a lexical category in the sense that it contains substantive meaning regarding the situation (Place & Manner) of an object. However, a caveat is in order here. The preposition also contains more abstract functional categorical features in the following way:
(18) Case & Prepositions
The specific features associated with the given functional / structure-class words and phrases will be more fully fleshed out in subsequent sections. What is important to understand here is the inter-relationship between lexical and functional categories--namely, that functional categories provide their lexical counter-parts with abstract grammatical material: (e.g., D-to- N, Aux/Modal-to-V, P-to-N, etc.). (Note that within our adopted binary notation of Case, [-Nom] (minus nominative), by default, automatically equates to [+Acc] Accusative Case). (19) Table: The "Four Parts-of-Speech" Categories--Lexical vs. Functional Status
Sentence (21) (22) (Lexical gibberish with syntactic meaning) These non-sense words could easily be syntactically slotted and thus spun into English counterpart parts-of-speech: toslaked (=saw), blevish (=pink), zimperstopen (=elephant), manikoning (=skating), etc.etc. Today, I saw a pink elephant skating down flower lane. The two sentences in (21a, b) are correct English expressions I have never spoken/heard before. The completely made-up sentence in (22) is also correct on pure syntactic grounds even though the individual words have no lexical sound-to-meaning relation in English. I doubt you have heard either of the sentences in (21a, b) uttered before in this exact wording--not to mention the completely gibberish sentence in (22) above. Nevertheless, I believe we can all agree that (21a, b) are indeed English sentences which project a certain meaning--albeit a meaning that might be better served in a fairy-tale novel. The fact that the gibberish in (22) can likewise be syntactically parsed as a possible English sentence immediately begs the following question: What is it exactly that allows one to process and perceive a given sentence? Just think about it--a sentence never heard/spoken before can instantly be comprehended without difficulty. Well, the magic of it all generally has to do with the sentence structure template and the ability of such a template to string certain words of word-classes together (Noun, Verb, Adjective, etc.) to form the subject & predicate. The fact that I can creatively generate these random sentences--seemingly stringing one word after another and with full comprehension on your part--is based on the fact that they are buttressed by an underlying common structure. It is owing to this structure that we come to an accepted comprehension--perhaps even more so the structure than the actual individual words that make-up its structure. So, returning to the issue of language providing a potentially infinite set of sentences (by which an infinite amount of words can be randomly shifted), we are in fact merely noting the myriad of possible word combinations: The term 'sentence' is more than just the total added value of the string of words put together, but rather something much more. (The sum is greater than the parts). A sentence is a very specific 'arrangement' of linguistic structure--the individual words simply serve to fill in the 'slots' (so to speak) of this structure. If--as earlier schools of thought might have had us believe (viz., Behaviorism)--grammar analyses relied on a simple collection of all possible sentence configurations, including all token sentence types with all possible word arrangements, etc. (sifting correct from incorrect types), just the task of simply sorting through the memorized data itself would have brought a break-down in our mental abilities. Such a heavy burden would leave very little computational room (mental capacity) for the actual subsequent processing of the arrived sentence. In fact, trying to process (parse) language word-by-word would put such a strain on short term memory that we would ultimately never be able to comprehend those more abstract or complex sentence. We can understand the sentences above because the overall structure is consistent with English sentence structure/grammar and the words themselves (although 'fairy-tale-like' or nonsensical ) are positioned in the appropriate 'slots'. If, on the other hand, we were to arrange the same words in (21a), say in any random order, then the slots would not map onto the accepted fixed structure and the sentence--even with the identical words--would not make any sense to us. Consider the new rendition of the 'elephant sentence' in (23) below now with an altered arrangement of the very same twelve words: (23) * Pink yellow and yesterday lane I flower a saw elephant down roller-skating. As we see, the expression makes no sense: even the gibberish but parsed sentence in (22) makes more sense to us. Having now convinced you ( I hope) that general Sentence and Phrase structure counts, let's now consider what the exact structure looks like--carrying over the labor of dividing and subdividing Sentence and Phrase Structure onto the remaining relevant sections of the text. (See Phrase for a full phrase analysis of the 'elephant sentence' (43)). The first thing we must understand is that the 'elephant sentence'--like any English sentence--is divided into two larger segments: (i) the subject and (ii) the predicate. Subject-Predicate
On the topic of "Sentence", what we also must recognize is that there is an order of different Sentence Types--all of which preserve the overall essence of "sentence" (per se, as a class type), but an order which also bears to light some fundamental differences regarding more subtle Features of structure. (Regarding features, one could say that the "selectional requirements" for a V(erb) <x> to select a specific type of Object <y> or a Non-object <z> in its predicate is determined by the verb's feature setting: e.g., [V: +/-object]). The sections below spell-out the major sentence types of English. The Latter sections (§§. 3, 4) will deal with matters regarding Complex Sentence, Sentence Structure and Movement.
The first type of sentence, and most basic type, is the Intransitive Sentence. This Type of sentence structure contains a certain class of Main Verb (MV) that doesn't necessarily need to have an accompanying Object (in the predicate) to serve as its complement. As we shall see in the latter examples, it's the Main Verb that delegates and projects whatever type of information is required for the predicate: e.g., whether or not one object, two objects or no objects are required. Consider the following examples below: (25) Table: Intransitive Sentences
What's important to note here with regards to the above sentences is that the mere projection of (i) a Subject and (ii) a Main Verb is sufficient in satisfying the requirements for a properly formed sentence: these requirements for a well formed sentence are governed by the semantic properties of the Intransitive Verb. Although we may wish at anytime to combine additional (Adverbial) material to the predicate--such as in the sentence e.g., "(Fish swim (fast/ under the sea))" etc.,--what is important to understand is that this additional material, here, taking the forms of an (i) Adverbial and (ii) Prepositional Phrase (respectively), is not an essential requirement of the verb. In other words, the verb "swim", being an Intransitive Verb, doesn't look leftward to its predicate (or complement) seeking assistance in maintaining the meaning of the expression--the verb is able to stand alone contributing 100% of its predicate (semantic) information directly back to the subject "Fish". Clearly, there is no aspect whatsoever of the semantics of "swim" which could refer to anything but the subject "Fish". We could expression this Intransitive property in the following logical expression: swim(Fish). When we say "Fish swim", 100% of all meaningful material is associated right back to the subject. In our example in (20a) above, "I (can) study [with the book]", the main verb "study" directly links back 100% to the subject without seeking out any additional support from an object. As stated above, the fact that we do have additional predicate material--in the form of a Prepositional Phrase (PP)--is superfluous to the nature of the verb "study", and is simply affording us with extra linguistic material that could otherwise be forgone. Of course, we could very well stop with the sentence "I study" without jeopardizing the verb's integrity. Another example of Intransitivity would be the verb "sleep"--as in John sleeps. No sense can be derived from *John sleeps Mary (=SVO). This malformed sentence arises because the verb 'sleep' must contribute 100% of its total meaning back onto its subject--i.e., the verb's meaning has absolutely nothing to say about any possible interceding object. However, note the well formed counter-example The general bedded the soldiers ('to bed' here meaning to supply bedding). The nature of the verb 'bed' requires an object--one receiving bedding. Consider how (25) above would look (and be diagrammed) without such extra material in its predicate:
These simplest sentence structures are considered "Intransitive" and consist of a Subject (DP) followed by a Predicate in which only a Main Verb Phrase (MVP) is required (followed by optional adverbial information). The name for such verbs which can stand alone in its predicate is technically termed "Intransitive". While such sentence types can rest with a prosaic Subject and Main Verb, they may optionally combine additional Adverbial material in their predicates. Consider how an Intransitive type structure would be Tree-Diagramed below:
1.2 Transitive Sentence: Copular Linking Verbs In contrast to what we have observed above with Intransitive type sentences, Transitive Copular sentences do require additional Adverbial information in the shape of either e.g., a Prepositional or an Adverbial/Adjectival Phrase (to is left in the predicate) in order to keep the meaning of the copular verb stable. Consider the following type of sentences below: (28) Table: Copular Transitive Sentences
By looking at the separate predicate constituents of the sentences above, you will see that they contain not only a main verb (=copular "Be"), but that they also must contain some kind of an obligatory Adverbial Phrase. What we mean by "Adverbial" is that the phrase supplies essential additional information to the verb (modifying the copular verb so to speak)--and so the label adverbial. The obligatory adverbial complement information is the following: outside, yesterday, on the desk, in the yard, a teacher, tired (respectively). Note that we are referring to the Prepositional Phrases (PP) here (on the desk, etc.) as being somewhat adverbial in nature. In the above sense, such PPs are not optional (as in Intransitive structures), but rather are required. In other words, if the adverbial complement were dropped, the copular transitive verb would not be stable enough to transmit meaning. Consider how such ill-formed copular structures would look without their adverbial complements: (*John is ø /Her job interviews were ø/ Mary's notebook must have been ø/ The part will be ø). As you quickly discover, the copular verb "Be" cannot stand alone in the predicate but must be supported by other adverbial material. Such supporting adverbial complements provide information regarding Place, Time or Manner/Mode: e.g., John is... in the class/late/tired (respectively). These copular verbs {Be--is/are/was/were} are called Linking Verbs due to the fact that they directly link-up the adverbial information to the subject: [Subject + Linking Verb]--i.e., 100% of the adverbial material directly reflects back to the Subject (Noun), making all such (linking) adverbial material in the predicate quasi-adjectival in nature. To a certain degree, linking verbs function in a quasi adjectival manner--describing or modifying the nominal (Noun) subject. When one says "Johni is a teacheri" (John=teacher), the DP-object [a teacher] directly co-indexes and refers back to the DP-subject [John] (as indicated by the subscript index 'i' )--much in the same way as a adjective reflects back onto the Noun its modifying. Such Linking verb constructs with co-indexed complement are often termed "Nominal Subject Complements". (See (31) below for structure). This is because the complement/object functions as a quasi-reflexive in conjunction with the nominal/subject. In fact, this modifying quality of the copular "Be" becomes even more apparent when copular structures are inverted as an Adjective Phrase: (29)
(NB. This verb form of nominal modification will become important when we come to examine how copular linking verbs of sense are modified not as verbs (par excellence), but rather as nouns). (See section (33) below). Consider the Copular "Be" tree structures below showing adverbial material in the complement slot positions:
Token Sentences (30e, c) above capture this generic Adverbial material--the object of the copular verb--by either projecting a more accurate Prepositional Phrase (ex. c) or a Determiner Phrase (ex. e) in the adverbial complement slots of the verb. Consider the revised complement phrasal projections below: 
Let's now turn to the issue of how other (copular) linking verbs function as quasi nouns--in terms of how they get modified. Verbs of sense are frequently used as linking verbs as in the following examples:
What is interesting about the sentences above is that the typical adverb modification of a verb--i.e., the grammatical structure of [Verb + Adverb], has become replaced by the structure [Verb + Adjective]. Note that the adverbial counterparts to the modification are incorrect: e.g., *The milk tastes sweetly is something you would never say on an intuitive basis. What we are suggesting here is that both copular "Be" verbs as well as Linking Verbs of the Senses take on a certain amount of Nominal (noun) qualities--so much so that when they enter into a modification structure, it is the adjective (a noun modifier) which wins out over the adverb (a verb modifier). The overall structure of modification suggests that copular as well as linking verbs get their lexical/substantive properties directly from the Subject (Noun)--hence, the notion of "linking": verbs that directly "link" the Subject to the Predicate (Object or Modifier). (See also (153) Copular/Main Verb "Be" vs. Auxiliary "Be" for further discussion).
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
.gif){kind=small}
