The Syntax and Morphology of English Verbs

Patterns that Matter

Joseph Embley Emonds

• Publisher: De Gruyter Mouton
• Year: 2022
• ISBN:
  9783110734416
• Edition: 1
• Topic: Langues et linguistique
• Subtopic: Linguistique

The book provides a detailed empirical approach to constructing grammatical analysis and theory, in particular the analysis of English verbs. It develops an integrated formal description of the English verbal system and offers several theoretical advances in the treatment of verbs that have escaped formulation until now.

1 English modals and be are not irregular verbs

1.1 English auxiliaries

The apparent complexity of the English system of verbal auxiliaries is well known and has been analyzed in many works on grammar, both theoretical and practical. To illustrate the variety of auxiliaries, let’s examine an artificially constructed but still simple exchange among four friends of a couple who are applying for jobs.

(1-1)

a

Are John and Mary maybe getting hired now?

b

John is already hired. Was Mary not being hired Monday? So they did get hired.

c

That’s wrong; John and Mary have not yet been hired.

d

But they will be hired soon, even if being hired does not seem easy and is not quick.

With the same main verb hired, these sentences contain twelve different words that are called auxiliaries: are, getting, is, was, did, get, have, been, will, be, being, and does. And all except get/getting and be/being appear to be unrelated in form to each other.¹

Traditional and structuralist treatments of the English verb go some way toward setting up an observationally adequate system, for example the “5 slot model” in (1-2) laid out in Quirk et al. (2004).

(1-2)

English predicates: (modal)-(perfect have)-(progressive be)-(passive be) – lexical verb

Nonetheless, these treatments end up stating that an English lexical verb such as hired in the examples above can be preceded by a string of up to four auxiliary verbs in the same clause (should have been getting hired), and that these verbs, being subject to item-particular restrictions, are, well, just irregular. Thus, the available descriptions are simply lists of unpredictable irregularities of form and combination, and make no clear line between which verbs are “lexical” and which are “auxiliary”.

The first decades of generative grammar saw significant progress in making formal sense of the English verb system. One important ingredient of this progress was the claim from the start (Chomsky 1957: ch. 7) that English auxiliaries and main verbs are not in the same syntactic category, even though some of the former may share some characteristic (later called a feature) of lexical verbs in some constructions.

This proposal was hotly contested soon after it was made, even by Chomsky’s most intrepid followers (for example, Ross 1969), and most generative work on verbs in the last thirty years seems to increasingly reject, or at least ignore, Chomsky’s break with tradition. As a result, current generative syntax can easily give the impression that it is content to continue to refine well-established tenets of traditional and structuralist grammar – that English auxiliary verbs are first and foremost verbs, but have irregular lexical properties (item-particular and peripheral) that partly obscure their membership in the category V.

The most intensely felt reaction to Chomsky’s original system and to subsequent proposals for maintaining and refining it seems to center on English modal auxiliaries, namely those which are invariant, whatever the grammatical number and person of a clausal subject. There is a clearly delineated class of modals that consists of the following: will, would, can, could, may, might, shall, should, must, ought, dare, need, and had better.²

This chapter explores the issue of whether modals are in the category V or not. A preponderance of authors in current linguistics, including generativists, assumes that they are. To answer this question, we first should have some contentful criteria as to what a verb is.

1.2 Diagnostics for the category verb

1.2.1 The place of verbs in clausal structure

In line with traditional grammar going back to Aristotle, verbs are thought to be words that play a central role in the construction of “complete sentences” or (when there are no embedded sentences) “clauses”. In turn, a clause is then composed of: (i) a subject, a potentially complex Noun Phrase currently widely notated as DP (“Determiner Phrase”), and (ii) a predicate formed around a Verb Phrase, a unit whose “main word” is a verb.³

From this perspective, a typical clausal structure thus looks like this, with the agreeing main verbs underlined:

(1-3)

Subjects + Predicates

[_CLAUSE [_DP [_D some] [_NP [_N trails] into the forest] [_VP really [_V look] steep]]

[_CLAUSE [_DP [_D no] [_NP [_N trail] into the forest]] [_VP really [_V looks] steep]]

[_CLAUSE [_DP [_D those] [_NP two [_N guys] you chose]] [_VP [_V try] to drive well]]

[_CLAUSE [_DP [_D that] [_NP one [_N guy] you chose]] [_VP [_V tries] to drive well]]

Thus one says that D is the head of DP, that N is the head of NP, and that V or VP is the head of the clause. Here are the coinciding defining properties of a head proposed in Harris (1957) and still current today:

(1-4)

Head Properties. A head of a phrase is a structurally obligatory position that selects other constituents in that phrase, and is selected by other heads outside that phrase.

The fact that a head is structurally obligatory does not mean that its position is always spelled out by an overt morpheme. The obligatory Noun positions e in the bracketed subject phrases in (1-5a) are all phonetically silent, as are the “gapped” obligatory Verb positions in (1-5b):

(1-5)

a

[Those e through the forest] are the most beautiful.

[Everybody else e in that neighborhood] seems to own a fur coat.

[Three or more e of them] are usually dangerous.

[e Who you marry] matters little to me.

b

Has your sister got a car or [your brother e a bicycle]?

John should clean the kitchen and [Mary e the bedrooms].

John is now a teacher and [Mary e a bus driver].

Although “head” is a central theoretical construct in syntax, there is no method of analyzing syntactic data based on a priori considerations such a...