Talk:Auxiliary verb/Archive 1

Archive 1

Archive 2

How did these come about?

I'm very curious as to why these verbs started to be used this way. In Middle and Old English one could say "I can music", so why did that change? When you think about it, many of these verbs are very illogical; to see the truth of this, just read this sentence: "Do you be hungry?" Pretty weird, eh? Well that's how weird "Do you want to dance" would sound if we weren't so used to the construction. I know I for one would love an explanation of how this bizarre phenomenon came about. Fuzzyblob 20:48, 28 March 2007 (UTC) no you see you may have misunderstood it would not say "do you be hungry?" it would say "would you be hungry?" —Preceding unsigned comment added by 98.195.184.10 (talk) 01:53, 16 June 2010 (UTC)

Go as a quasi-auxiliary

In the sentences "I will go fishing" and "We went shopping," is go a quasi-auxiliary, or something else which I can't guess at? Fishal 18:47, 18 April 2007 (UTC)

I may be mistaken about this, but I believe that both fishing and shopping are functioning as gerund nouns in their respective sentences. In both of these sentences, the speakers are merely talking about the acts of fishing and shopping; nobody is actually fishing or shopping in the sentences. —Preceding unsigned comment added by 209.204.84.2 (talk) 13:24, 25 April 2008 (UTC)

aux verb ase

he as gone ( can tell whule talking) —Preceding unsigned comment added by 210.212.201.50 (talk) 13:16, 4 September 2007 (UTC)

Do you mean "he has gone"? —Preceding unsigned comment added by Mrboire (talk • contribs) 13:29, 28 April 2009 (UTC)

Could as past tense of Can

I might joke when somebody asks; 'Could I help you?'. My joking reponse might be 'Yes you could, but you can't now'. Is this an example of could being interpreted as past tense?

The article says 'Could I get you something?' clearly is not expressing past time. In my opinion, it is expressing the past time, but (depending on context) is likely to be understood to mean the same as 'Can I help you'

hrf 10:57, 7 September 2007 (UTC) (England)

English in general does allow the usual past tense to be used with a sense of remoteness in some cases (consider a shopkeeper asking, "Did you need help finding anything, sir?"), but I think it's pushing it to claim that there's "past time" here. —Ruakh_TALK 18:16, 7 September 2007 (UTC)

Need, Dare

there are modal verbs, is not there?! But why they did not include in modal verbs schedule. The negation form is 'dare not' and 'need not'. I have always thought that this two verbs are auxiliary. Weird.

Dare is a semi-modal verb. The speaker can choose whether to use the auxiliary "to" when forming negative and interrogative sentences. For example, "I don't dare (to) go" and "I dare not go" are both correct. Similarly "Dare you go?" and "Do you dare (to) go?" are both correct.

Need: (modal verb) To be obliged or required to.

You need not go if you don't want to.

Modal verb talks about "dare" at least. Not sure as to the reasoning of non-inclusion in this article. Remember Wikipedia is always a work in progress!--Never give up! Never surrender! (talk) 13:28, 28 April 2009 (UTC)

Is a linking verb really an auxiliary verb?

The article lists linking verbs as a synonym for auxiliary verbs, but iirc linking verbs were from sentences of the form:

• subject linking-verb adjective
• subject linking-verb noun

Such as

• He is happy.
• They are tired.

I am not a grammar expert but I'd like it if someone who is can verify whether or not linking verb should be listed as a synonym for auxiliary verb.66.32.173.12 (talk) 04:42, 26 September 2009 (UTC)

Since linking verbs refers to copula verb and not to an auxiliary verb, and no one has stated otherwise, I'm removing it from the list of synonyms for linking verb. Please revert if you are sure this is in error. 66.32.195.121 (talk) 17:36, 8 October 2009 (UTC)

Expert needed

The lead says that the two main auxiliaries are "to have" and "to be", then the section Functions of the English auxiliary verb only shows the functions of "to be". "Need" and "dare" are not listed as auxiliaries. No references to be seen. A separate article has a list of English auxiliary verbs which is shot to pieces and should probably be stripped of considerable rubbish and incorporated here. Please someone help clean up? Thanks. 125.175.98.178 (talk) 13:05, 28 May 2010 (UTC)

I am not an "expert" but I do know that much of the article is incorrect.

• The copula, "to be", is not an auxiliary verb. It either stands on its own or
  • The passive voice in English is formed by the verb "to be" and the past participle.
  • The progressive tenses are formed by the verb "to be" and the present participle.
• The verb "to have" is not an auxiliary verb. It either stands on its own to show possession or
  • The present perfect is formed by the verb "to have" in the present tense and the past participle.
  • The past perfect is formed by the verb "to have" in the past tense and the past participle.

A true grammarian or, at least, a member of the League of Copy Editors needs to take this article and rewrite it. JimCubb (talk) 23:44, 28 August 2010 (UTC)

"Know"? Re "The verb 'to have' is not an auxiliary verb": Crystal, David (1995), The Cambridge Encyclopedia of the English Language, Cambridge: Cambridge University Press, ISBN 9780521401791, pp. 207, 212; and http://www.merriam-webster.com/dictionary/have. As for "true grammarian", see Crystal 1995, p. 194, and much more importantly, Pinker, Steven (2000), The Language Instinct, New York: Perennial Classics, ISBN 9780060958336, pp. 382-418. — ¾-10 02:16, 1 September 2010 (UTC)

The article is correct in this regard. Both of the above non-stand-alone examples of "to be" usage, and both non-stand-alone examples of "to have" usage, are in fact examples of these verbs being used as auxiliaries. 75.183.96.242 (talk) 17:17, 5 October 2010 (UTC)

Distinguishing auxiliary verbs from adverbs

In the new section on Mandarin Chinese, I have given Li and Thompson's two criteria for distinguishing an auxiliary verb from an adverb. The second criterion is language-specific, but I wonder if the first one is universal, and if indeed it should be stated in the intro of the article: auxiliary verbs are distinguished from adverbs in that yes-no questions can be answered with subject + auxiliary but not with subject + adverb. E.g., in English "Are you definitely coming?" can be answered by "I am" but not by "I definitely".

Does anyone know if this is a universal necessary criterion for classifying a word as an auxiliary verb? 75.183.96.242 (talk) 17:27, 5 October 2010 (UTC)

Dubious about the whole article

I think this article does a poor job of describing English auxiliaries. I have made no effort to look at the material about other languages. There is a major distinction between the "have" and "be" that build the aspect and voice and the modals - "can" etc. Culicover in his "Syntax" argues that the modals are not part of the verb phrase. I disagree - but I have not published my disagreement so my version is not suitable for Wikipedia. I suggest somebody go get a good modern reference grammar of English and describe what it says DKleinecke (talk) 17:52, 6 November 2010 (UTC)

One problem is that we need to develop explanations of how linguistic science differs from traditional pedagogy (e.g., EFL and K12) throughout WP's language coverage

[Moved from User talk namespace for future pondering by others:]

Hey,

I appreciate that the emphatic is taught in ESL and other such texts as a use of 'do'. Unfortunately though it's wholly incorrect. The emphatic in English is expressed almost purely though tone and stress. The only role of 'do' (when used as an auxiliary) is as an aspectual auxiliary (an auxiliary used to show aspect) signifying the non-durational aspects (versus 'be' which is used to signify the durational aspect). The aspectual auxiliary 'do' when used in an affirmative statement in the present and past tenses in non-perfected constructions may be omitted with its ending appended onto the content verb as a declined form. For example do + walk = walk (zero ending), does + walk = walks, did + walk = walked. This linguistic shorthand is only allowed in this one type of sentence in these two tenses. Otherwise the auxiliary and content verb must be kept separate (which is why negatives and questions have 'do' all the time and affirmative statements only have it sometimes).

The emphatic in English is simply expressed by putting emphasis on the desired element. That emphasis may be applied to the subject, content verb, auxiliary, or any other element of the utterance as the speaker deems most effective. Emphasizing the auxiliary is one of the more common methods. To those who do not understand the role of aspectual auxiliaries and recognize 'do' as one, it may appear that do is present in emphatic situations but missing in others (assuming 'do' is the emphatic) whereas in fact, the emphatic may be expressed with or without the auxiliary, AND the auxiliary do was there all along, just "hidden" in the content verb.

This is why I'd removed the sections on "dummy verb" and "emphatic" because as you see above, there is no such thing as a dummy verb and the emphatic has nothing more to do with 'do' than it does with any other word in a sentence. This is a case of just because something is taught does not make it correct. Finally, this is an encyclopedia so the information provided should be based on linguistic understanding and not misguided teaching pedagogy.

I'll leave the article as is for now (the whole thing needs a rewrite really because so much is incorrect in it) but I encourage you to reverse your undo based on the above. Also just FYI I am a linguist specializing in English and also write ESL books as well so there's no need to request a further 'expert opinion', you've just gotten it. (not trying to be a smartass, but I've had too many debates on wikipedia about linguistics only to find that the person debating the topic wasn't a linguist and assumed I wasn't one either).

Thanks Drew.ward (talk) 02:47, 21 April 2011 (UTC)

Hi—no problem, and in fact, I apologize to you. I actually did suspect that this might be a case of the very theme that you explicated above, where traditional language epistemology is incorrect versus linguistic science. Really I should have asked you about that before restoring the paragraph, given that I had this suspicion, rather than restoring it first and then forcing you to defend your edit. My only excuse for restoring first and asking questions later, as I analyze my behavior, is that my WP:AGF-vs-expediency ratio when watchlisting occasionally dips too low toward the latter on corner cases (and the epistemological problem of language instruction creates some watchlisting-heuristics corner cases on language articles). In other words, what you said is exactly correct—"the person debating the topic wasn't a linguist and assumed I wasn't one either". I apologize for that, as the last thing I want to do is to discourage trained linguists from improving WP's articles on language topics! I'll go self-revert in a moment. Regarding the two epistemologies (conventional [flawed] schoolroom pedagogy versus linguistic science), I think that WP eventually needs to mention both, and then explain how the former was wrong, corrected scientifically by the latter, but the former is still being used for pedagogy because it's speciously attractive. Possibly even "useful", if incorrect theory that yields decent empirical results is truly as "useful" (net positive vs net negative) as it seems on the surface. Which it may not be in fact!! But for humans to let go of it, they'll need Wikipedia to explain patiently why it's wrong, and then they'll need some years of slowly letting go of it before it gets dropped entirely. This, of course, is not the sole responsibility of editors like you; just the dream of what's ideally needed, to be pursued however the resources can allow. End of digression—thanks for your patience, and sorry I tested it in this case! Regards, — ¾-10 12:09, 21 April 2011 (UTC)

Sorry I missed this discussion back in April. Of course there is such a thing as dummy "do": e.g., in "Do you go?", "do" plays no role here other than as a place-filler for an absent auxiliary that needs to be there to convert to question form. And of course there is such a thing as emphatic "do": e.g., "He does seem confused" expresses more emphasis on "seem" than does "He seems confused". And no, "do" is not an aspectual auxiliary for the non-durational aspects: e.g., in "He is not a very social person. But he does belong to several clubs", "does belong", like "belong" in isolation, is durational; the insertion of "does" does not alter the aspect. Duoduoduo (talk) 17:24, 3 June 2011 (UTC)

Duo it very much is the aspectual auxiliary for the non-durational aspects in English. The verb BE is the exception, but then again it's the exception for almost everything. The reason you're confused by the use of DO with a verb like belong is because the aktionart of 'belong' is durational. Aspect is only used to emphasize or override aktionsart. When a verb having durational aktionsart is used in a non-durational aspect, the effect is still durational. Aktionsart is the most commonly ignored yet probably most important feature when determining the information conveyed in an utterance.Drew.ward (talk) 00:34, 4 June 2011 (UTC)

Nope, I'm not confused. You say "When a verb having durational aktionsart is used in a non-durational aspect, the effect is still durational." That just doesn't make any sense: if the effect is "still durational", the verb is not being "used in a nondurational aspect". If it's durational in effect, it's durational in aspect, because aspect is the effect being conveyed. So there is nothing non-durational about "do belong", because as you say the phrase has durational effect.

Here's another, non-aktionsart example: "He goes there frequently" is in habitual (hence durational) aspect, even though "goes" is not innately habitual. And "He does go there frequently" is still habitual, and hence still durational; the auxiliary "do" does not impart non-durational effect, so it is not a non-durational-aspect auxiliary. Instead, in this example, "does" adds emphasis. Duoduoduo (talk) 17:29, 4 June 2011 (UTC)

Actually you are a bit confused. There is no such thing as a "non-aktionsart example". All verbs have an aktionsart that is inherent in their meaning. Go does have a durational aktionsart. Aspect is not the whole picture as you're seeing it. Aspect is the use of structure to define certain attributes of the temporal nature of the (whole) verb. The primary attribute expressed via aspect is duration. Perfection is aspect's counterpart in that it's the use of structure to express completeness of the verb (the other half of aktionsart). In English duration and completeness trump a lack of either. So in a non-durational aspect (such as your habitual example) the aspect is non-durational but the aktionsart of the inclusive verb 'go' is durational so the overall temporal nature of the utterance is durational.

All the non-durational aspects in English use DO as their aspectual auxiliary (including the habitual). In the same way the durational aspectual auxiliary is BE and the perfecting auxiliary is HAVE. Drew.ward (talk) 21:04, 4 June 2011 (UTC)

You say "Aspect is the use of structure to define certain attributes of the temporal nature of the (whole) verb." Actually aspect is the temporal nature of the action that is conveyed by a clause and its verb--whether it's conveyed lexically or by grammatical structure. You say "The primary attribute expressed via aspect is duration." "Primary" is a matter of opinion there--I would view the nature of duration (habitual, stative, progressive) as being just as important.

You call the habitual aspect non-durational -- that's not right: It's durational (see e.g. Comrie's Aspect).

Consider this example: "blink" is innately non-durational. The verb in "He blinked" conveys non-durational aspect. But "He blinked repeatedly" conveys habitual aspect (via the adverb) and hence duration, and "He did blink repeatedly" conveys habitual aspect (again via the adverb) and emphasis on the verbal phrase (via "did"). "Did" here has nothing to do with non-durationality--it conveys emphasis.

Another example is "He did know that". "Know" is innately stative, and thus durational. The inclusion of "did" does not convey any non-durationality; it conveys emphasis.

You say "Perfection is ... the use of structure to express completeness". Here you seem to use "perfection" in the sense of the perfective aspect. But later you say "the perfecting auxiliary is HAVE". Here again you seem to use "perfecting" to refer to the opposite of durational (since you contrast it to "be"); but actually "have" is used for the perfect tense-aspect combination, not the perfective aspect. They are entirely different things. "Have" does not convey perfective aspect -- quite the opposite; it conveys at least two points in time (a point in the past, and now) ("I have gone to Paris") and often conveys many points in time ("I have studied for five hours now"). Duoduoduo (talk) 23:40, 4 June 2011 (UTC)

Actually I didn't say perfection was the opposite of aspect at all. Check what I've written above. Aspect and perfection are both attributes of the utterance versus aktionsart which is an attribute of the verb itself. Completeness (telic/atelic) and duration (durative/adurative) together make up the aktionsart of a verb. States, accomplishments, achievements, etc can all be defined as a combination of these two attributes at their core. In languages like English where aktionsart is a strongly governing force, the aktionsart of the verb by default, sets that combination of duration and completeness for the utterance in which it appears.

Aspect and perfection can be used to establish duration and completeness of the verb at a morphosyntactic level (as an attribute of the utterance versus a lexical attribute of the verb). English has multiple non-durational aspects and a single durational aspect. Don't mistake this to mean that utterances with a non-durational aspect do not express duration. Non-durational aspects are simply those that do not add duration. If the inclusive verb has a durational aktionsart then that utterance shall be durational because duration is already set by the verb and non-durational aspects are neutral that they don't change the durational nature as established by aktionsart. Durational aspects however specifically express duration through the structure of the utterance.

Thus a verb with an adurative aktionsart when used in a non-durational aspect will still lack duration. An adurative verb used in a durational aspect will have it's adurative aktionsart overridden by the aspect to add duration where none was before. A durative verb used in a non-durational aspect retains it's durative qualities. Using a durational aspect with an already durative verb either emphasizes or amplifies that verbs original durative qualities.

Aspect and aktionsart are two different things. Likewise, perfection is a means of completing the verb at the structural level. It works the same way as aspect by either being neutral, overriding, or emphasizing the existing telicity of the verb as defined by it's aktionsart.

Perfection is separate from aspect (and is itself not an aspect) because any verb can be perfected regardless of aspect.

Calling perfection, or more specific the erroneous nomenclature 'the perfect' an aspect makes no sense because verbs in any aspect can be perfected. Also perfective/imperfective aspects are something entirely different from perfection. It's also important not to conflate tense with perfection and aspect as you've done above as tense can only be determined by knowing the aspect, perfection, mood, and aktionsart of the utterance first.

In English, DO is the non-durational aspectual auxiliary. BE is the durational aspectual auxiliary. HAVE is the perfecting auxiliary. Drew.ward (talk) 00:32, 5 June 2011 (UTC)

Recent edits by Capnprep

I have reverted the bulk of your latest revision but retained the changes you'd made to the section on progressive aspect.

I have however undone the rest as your version ignores the use of DO as an aspectual auxiliary for the simple aspect just as BE is used as aspectual auxiliary for the progressive aspect. Your assertion of DO as an emphatic modal is also false and should not be included. This is a very common misconception and one that fails even the lightest linguistic scrutiny. Please do not revert these changes without discussion them here so that we may resolve any disagreements.Drew.ward (talk) 07:34, 5 May 2012 (UTC)

Having read discussions you have had with other editors, I am not optimistic about resolving our disagreements over the content of this and other articles. The main problem here is the total lack of sources in this section, which means that any statement can be removed at any time, or preferably replaced with sourced information. Since your edits are not sourced, they are subject to removal and should not be restored until this problem is addressed. The previous content was also unsourced, but I choose to restore most of it because:

1. Other things being equal, long-standing unsourced material is preferable to newly-introduced unsourced material, because it is more likely to represent the consensus of editors that watch the page.
2. Other things are not equal in this case: The previous version was more accessible, more intelligible, and more correct than your proposed changes.

In detail:

• In the lead, since the auxiliary verb do is not used with infinitival to, I am changing to do to do (and to be to be for consistency).
• I agree that the main function of do was not treated in the previous version. For now I will simply add a link to Do-support and a summary statement. Your section on "Simple aspect" is dubious. If it simply represents what you believe, it has no place here. If it is not OR, source it.
• As you say yourself, the use of do for emphasis is recognized in most grammars. It may be — although I doubt it — that "the lightest linguistic scrutiny" suffices to invalidate this idea. But it can't be your own scrutiny; it has to be scrutiny reported in a reliable source.

I will continue reverting your edits as long as their content remains unverifiable. CapnPrep (talk) 12:27, 5 May 2012 (UTC)

Modal verbs with no gerund

I am not an expert by any means, but I was wondering, in the section, List of auxiliary verbs in English the final sentence says that modal verbs are defective in that they, "cannot be inflected, nor do they appear as infinitives or participles" and I was just wondering whether we can add gerund to that list? There is no coulding or shalling, at least not in English, but I'm not sure if this is diagnostic of modal verbs or just a coincidence. Cottonshirtτ 08:05, 16 July 2012 (UTC)

English verbs have two forms which are called participles, the past ("walked") and the present ("walking"), which has also been called the "gerund-participle". The distinction between the present participle and the gerund in English is without basis and is a carry-over from Latin grammar. There is a brief discussion and a reference to "The Cambridge Grammar of the English Language" (Pullum and Huddleston) at [1]. Count Truthstein (talk) 12:52, 16 July 2012 (UTC)

I can agree with this point. I find the distinction between present participles and gerunds dubious. I have added "gerund" to the relevant spot, however, since gerunds are indeed widely acknowledged in the literature. If someone wants to remove it, however, I will not object.

The distinction is quite simple: a gerund is a verb-derived noun while a present participle is a verb form that is very much still a verb and cannot function as a noun. The similarity is purely structural and not functional (and for that matter is structurally identical with some infinitives and if you'd want to separate participle-derived adjectives, then there's an additional structurally identical yet functionally separate -ing).Drew.ward (talk) 15:18, 16 July 2012 (UTC)

Where is all this coming from?

Woah! The changes that have been made to this article since I last paid attention to it (within the past 2 months) are ridiculous! It reads as if it's been written by someone who doesn't even understand what auxiliaries are or what they do. It's now full of generalisations, assumptions, and references to ideas that are neither accepted nor even remotely established.

Where is all of this coming from?

I've just removed content listing only a specific list of auxiliary verbs (no), a line in the introduction that treats them as "nonsemantic", and an entire section on some crazy idea called "light verbs". What's next, diet gerunds?Drew.ward (talk) 13:07, 16 July 2012 (UTC)

Tone of comments

Drew,

I have just taken a look at your specific comments above. You are characterizing my points as "ridiculous". Please refrain from using such wording in this exchange. You are increasing my willingness to fight your edits and comments at every opportunity.

Three sources are now named that employ the not-diagnostic for identifying auxiliary verbs. The following two counterexamples, one of which you produce above, are lexicalized:

a. Ask not what your country can do for you...

b. I kid you not.

These are lexicalized expressions, set combinations that are not used in productive speech, but rather they are pulled out of the speaker's mental lexicon as complete expressions. They do not challenge the validity of the negation diagnostic.

Concerning dare, need not, and ought to, the article mentions that these expressions can be viewed as auxiliaries. There occurrence as auxiliaries is, I believe, relatively seldom in everyday productive speech. For instance *Ought I help? or even ?Ought I to help? are certainly dubious formulations to varying degrees. But let me emphasize, the text acknowledges that these expressions can be classified as auxiliaries! A compromise here would be to go ahead and add them to the list, but then to add a statement to the effect that their status as fully productive auxiliaries in everyday speech can be debated. How's that for a compromise? --Tjo3ya (talk) 20:38, 16 July 2012 (UTC)

Diagnostics for identifying auxiliary verbs

I've removed the information regarding negation and limited declension as it is incorrect. Consider for example NEED as auxiliary (I need not go. She needs not go.) Also, while not as often used, the examples given in which not follows a non-auxiliary verb are linguistically well-formed and can express valid ideas (say for instance if someone doesn't eat spaghetti or, someone who eats not spaghetti but lasagna). The addition of this section is a neat idea and I commend the effort, but such discussions usually rely on vast generalisations and assumptions -- dangerous ground in linguistics and grammar.Drew.ward (talk) 00:40, 16 July 2012 (UTC)

Drew, You are correct that the negation diagnostic was presented in a manner that was glossing over some caveats. I think the diagnostic is, however, quite valid. The verb need can indeed be viewed as an auxiliary, but for me it sounds a bit stilted when it is used as an auxiliary. It is about like ought and dare, which can also be viewed as auxiliaries, but sound stilted (at least to my ear), e.g. Dare I try it?. Nevertheless, the text further above now mentions need as a potential auxiliary. And as a side note, your sentence She needs not go is unacceptable English; it should be She need not go. When need functions as an auxiliary, it cannot be inflected. Tjo3ya (talk) 04:28, 16 July 2012 (UTC)

I suppose my response below wasn't what you were looking for so I will add more: "The verb need can indeed be viewed as an auxiliary, but for me it sounds a bit stilted when it is used as an auxiliary. It is about like ought and dare, which can also be viewed as auxiliaries, but sound stilted (at least to my ear)"

Really?? You are basing your justification for including a topic you acknowledge to be flawed based on your opinion that you don't like the way the words SOUND?? Do I even need to type anymore here?Drew.ward (talk) 15:30, 16 July 2012 (UTC)

Further, need is not a "potential auxiliary". Need, dare, and ought have been used as auxiliaries in the English language longer than want has. This sort of argument and wording is absolutely ridiculous and baseless.Drew.ward (talk) 15:30, 16 July 2012 (UTC)

Your point about negation being able to be follow a full verb can be valid in cases where the full verb precedes another verb, e.g.

a. We tried not to laugh.

b. She began not to help.Tjo3ya (talk) 04:28, 16 July 2012 (UTC)

Try and begin in your examples are the main verbs of the utterance with DO being the auxiliary in each. "not to laugh" and "not to help" are infinitive phrases acting as the objects of each of those verbs.Drew.ward (talk) 15:30, 16 July 2012 (UTC)

As the negation diagnostic was presented (before you removed it), it was formulated in terms of postdependents. That is, the negation not cannot appear as a postdependent of a finite verb unless that finite verb is an auxiliary. Formulated in this manner, examples like the two just given do not contradict the diagnostic. The negation in these two sentences is a predependent of the to-infinitive; it is a not a postdependent of the finite verb. It is negating the following verb, not the preceding one. I do not think it is worth trying to explain this caveat, however, since doing so would occupy too much space and challenge too many readers.Tjo3ya (talk) 04:28, 16 July 2012 (UTC)

There are plenty more examples out there. Here's one terribly obscure one: "Ask not what your country can do for you..."Drew.ward (talk) 15:30, 16 July 2012 (UTC)

Your example someone who eats not spaghetti, but rather lasagna is consistent with the same point. The negation is a predependent of spaghetti; it is not a postdependent of the finite verb eats.

Further, the negation diagnostic for identifying auxiliaries is not something I made up, but it is employed in syntax and grammar textbooks that discuss auxiliary verbs. A couple of them were listed in the footnote you cut out.

I am going to reinstate the text you cut, but I will first wait for your response here. --Tjo3ya (talk) 04:28, 16 July 2012 (UTC)

"You are correct that the negation diagnostic was presented in a manner that was glossing over some caveats. I think the diagnostic is, however, quite valid." If a diagnostic needs caveats or a rule needs exception, then neither are worth including in an article.Drew.ward (talk) 13:08, 16 July 2012 (UTC)

This seems rather strong. The partial behaviour of dare and need as auxiliaries in contemporary English, with some definite dialect differences, is well-established in the literature and in actual use. It's only partial but that doesn't alter the fact that in a specific usage both words can function as auxiliaries. "Daren't you ...?" is a relative common construction (try a Google search for this phrase but note that the count doesn't reflect this precise grammatical form). What is dare in this construction? Answer: "an auxiliary". To say that because exceptions exist a rule should not be in a Wikipedia article is not a sensible position to take, in my view.

(On the dialect differences, it's interesting to me to see that "you daren't, dare you?" which is more natural to me gets only 10,000-odd Google hits compared to over 700,000 for "you don't dare, do you?" which is definitely less natural to me. But I'm a 60+ speaker of British English, which may explain why my idiolect is not common. Perhaps in another generation or so, dare and need will have been regularized as non-auxiliaries. But it hasn't happened yet and the current position should be acknowledged in Wikipedia.) Peter coxhead (talk) 14:28, 14 September 2012 (UTC)

Hello Peter,

Thanks for your comment. I'm not quite sure where you come down on the issue, however. Is there something about how the article is currently formulated that seems inaccurate to you? If yes, what should be changed?

Concerning your specific example "Daren't you...?", that expression is not in my active (North American) vocabulary. The expression Darned you! is, however. Are we talking about the same thing with these examples? If we are, then I would argue that we are dealing with a particular lexicalized form, which would mean that dare really cannot be judged to be a fully productive auxiliary verb in modern (North American) English. --Tjo3ya (talk) 23:13, 14 September 2012 (UTC)

Dare may or may not be a fully productive auxiliary verb in modern North American English; I wouldn't know. But it certainly is in other dialects of modern English. So the article should at least mention that need and dare have uses in which they function as auxiliaries. This can be sourced to, e.g., Radford, Andrew (1997). Syntactic theory and the structure of English. Cambridge University Press. ISBN 0-521-47125-7. p. 494: "AUX/Auxiliary A term used to describe items such as will/would/can/could/shall/should/may/might/must/ought and some uses of have/be/do/need/dare."

There is no dispute about that. That ought, dare, and need can be classified as auxiliaries is acknowledged in the article. Take another look. --Tjo3ya (talk) 17:47, 15 September 2012 (UTC)

What's wrong with the article as it stands is that it implies that a lexical item either is or is not an auxiliary rather than that in a given usage/context an item can be identified as an auxiliary.

This point easily accommodated. I'll add a sentence or two to the article.

In the sentence "You daren't do it, dare you?" which is perfectly acceptable in some dialects of English, the lexeme dare functions as an auxiliary using two tests (negation, question tag). In the sentence, "You don't dare do it, do you?" which is perfectly acceptable in some (not necessarily different) dialects of English, dare functions as a verb. The lexical item dare doesn't change; its use in the sentence does.

I'll add a sentence about dialects at the correct spot. No problem there. --Tjo3ya (talk) 17:47, 15 September 2012 (UTC)

The article also omits to mention uses in which have and be function as verbs but show behaviour otherwise diagnostic of auxiliaries, e.g. "Are you happy today?", "Isn't he John Smith?",

Please take another look at the article. The copula is listed as an auxiliary in two spots. --Tjo3ya (talk) 17:47, 15 September 2012 (UTC)

"She has a lot of clothes, hasn't she?", "Haven't you any food with you?". Like the use of need and dare as auxiliaries, treating have in this way shows dialect differences and may be diminishing. (Younger people tend to insert "got", as in "She's got a lot of clothes, hasn't she?" which can be analysed as have = auxiliary, get = verb although there are problems with this.)

I do not disagree about the validity of such data. I question, however, whether such data should be accommodated in the article. That strikes me as going into a level of detail that is not helpful, especially since I think a large number of speakers would avoid those formulations. But even if the tag you give here is produced, it does not contradict any statements in the article. The fact that such tags are formed with auxiliaries is in no way controversial. --Tjo3ya (talk) 17:47, 15 September 2012 (UTC)

However, this is perfectly normal usage of be in all dialects.

My underlying points are:

• The article does not display WP:NPOV, being biassed to North American dialects.

The bias you sense is minor, I think. If it something that really detracts from the article, I suggest drafting an additional section that might appear with the title:"Variation across dialects of English". --Tjo3ya (talk) 17:47, 15 September 2012 (UTC)

• The article over-simplifies the actual grammar of English, misleading readers.

How does it mislead readers? Concrete example? --Tjo3ya (talk) 17:47, 15 September 2012 (UTC)

I come to this from computational linguistics, where it's well-understood that lexemes in English cannot be assigned rigidly fixed categories in parsing actual text and speech, since one of the striking features of English is its flexibility in this respect. Thus during the Olympics many commentators noticed the use of medal as a verb as in She medalled today, which to most people was a novel use. For people and computers to successfully parse English, they must be able to deduce the category of a lexeme from its use, not from a dictionary entry, although this can provide the starting point by giving the most common category. Peter coxhead (talk) 16:46, 15 September 2012 (UTC)

That's obvious. How does the comment pertain to the article? Perhaps a sentence stating that many of the categories can also function as full or light verbs would take care of the issue. --Tjo3ya (talk) 17:47, 15 September 2012 (UTC)

Ok, just commenting on the article.

• It claims that the copula be is an auxiliary. Please provide a source for this. The sources I have to hand don't support this view, but rather say that in "Is she happy?" the copula be is NOT an auxiliary, but that English speakers extend its behaviour when an auxiliary to this construction, as they do in the tag in "She has a lot of clothes, hasn't she?" in those dialects which allow this construction. The article nowhere says this.

• Although the opening sentence says (correctly in my view) that auxiliary refers to the use, later sections are written as though words simply are or are not auxiliaries. This is not true of be in any dialect according to the view noted above, nor is it true of have in its possessive sense, need or dare in some dialects.

A comment made elsewhere on this page seems relevant: the article needs to explain what some referenced textbooks say about auxiliaries, not what any of us think. I'm as opinionated about this as anyone else! Peter coxhead (talk) 20:03, 15 September 2012 (UTC)

Peter, check the sentence in the section "List of auxiliaries". There it is stated that the verbs have and do can also function as main verbs or light verbs. That sentence addresses much of the problem that you have with the point about lexical items. Should I add a second statement to that effect, perhaps in the introduction? If that is your main concern, I think it is easily accommodated.

Your point about the status of the copula be is more problematic. I may or may not be hard-pressed to find a source that states that the copula is an auxiliary. In my personal opinion however, there is no reason NOT to classify the copula as an auxiliary. I think it passes all the criteria for defining an auxiliary (it lacks full semantic content; it is not a separate predicate and hence cannot be negated, it allows subject-aux inversion, it allows the appearance of not as a postdependent, many languages lack its equivalent). But if you are now going to press the issue, I will concede the point. Perhaps the copula can be removed from the table listing the auxiliaries. A clarifying statement could be added in the paragraph after the table to the effect that the copula, while it technically is not considered an auxiliary, nevertheless behaves like an auxiliary according to the criteria discussed for identifying auxiliaries (subject-auxiliary inverstion, appearance of not, verb phrase ellipsis).

Concerning sourcing the article better, I'm definitely in favor of that. I you have the sources available and see where these sources can be relevantly cited, I'm in favor of including more citations. I suggest adding the sources yourself. I can come in later and ensure that the format is all consistent. --Tjo3ya (talk) 20:46, 15 September 2012 (UTC)

I'm extremely reluctant to edit this article, because it's an area where I have strong opinions, and it seems to me that this is one of the problems with the article: it's been edited based on editors' opinions and not on what sources say. (I mainly edit plant articles, where I am an "informed amateur", with no pretensions to professional opinions.) The real point is that it should not matter in the slightest what you or I think. WP articles should be encyclopaedic, reflecting accurately what secondary or tertiary sources say. If some sources classify the copula as an auxiliary, then say so, with a source. If some sources don't classify it as an auxiliary, then say this too, with a source.

[As it happens, I can't find any source which considers the copula to be an auxiliary, but there again the sources available to me are not representative: my professional interest (before I retired) was in phrase structure grammars and their subsequent derivatives. In such grammars, the key point about an auxiliary is that it occurs as the head of an Auxiliary Phrase which also contains a main verb, actually or elliptically. Since this is not true of the copula, in such grammars it can't be an auxiliary.]

I'm not one of those editors who likes to go through articles tagging them for problems, but as an exercise to point out the problems I'll do so in this case. Peter coxhead (talk) 09:01, 16 September 2012 (UTC)

Peter, I agree with your emphasis on sourcing and welcome the scrutiny. I disagree with your stance that experts should not be contributing to the articles in their field; informed amateurs can be just as biased as they strive to reproduce the analyses and arguments of the experts they emulate.

Very true! But, in my experience, it's often easier for informer amateurs to be at least less biassed. Peter coxhead (talk) 09:35, 17 September 2012 (UTC)

Jurafsky and Martin (2000:320) write "English auxiliaries include the copula be...". Some other textbooks, I'm finding, define the copula be merely as a "linking verb", without stating whether a linking verb should be viewed as a full verb or as an auxiliary. Interestingly and tellingly, the other copular verbs, i.e. copular verbs other than be, are clearly NOT auxiliaries, e.g. *Taste that good?, *It tastes not good, *Became you angry?, *You became not angry, etc.

You state that your sources make it clear that the copula be is not an auxiliary. Would you please take the time to list one or two of the sources in full and the exact page numbers where this claim is made. I want to incorporate that claim into the article.

Your statement about VPs can actually be construed as support for the stance that the copula be is an auxiliary. Notice that like the other auxiliaries, copula be licenses verb phrase ellipsis: She is friendly, and he is too. In this example, the predicative adjective friendly is elided just like a VP would be.

I am now going to add some sources to the article and in the process, I will be removing some or all of the recent tags you put in. --Tjo3ya (talk) 22:29, 16 September 2012 (UTC)

DO the various sources that both of you are drawing from specify whether they are classifying be when functioning as a copular as an auxiliary or whether they are asserting that be in general is or is not an auxiliary? In other words do they separate the two or treat all appearances of be regardless of role as copular? These could be two very different directions depending on how they are classifying be. Some authors seem to classify only verbs that can function only as auxiliaries (cannot be used as the content verb of a construction) as auxiliaries with examples like must, can, shall, etc being auxiliaries but have, do, want, etc not being auxiliaries because they can be have non-auxiliary functions.Drew.ward (talk) 23:11, 16 September 2012 (UTC)

Hi Drew, The dominate practice seems to be to draw a vague distinction between auxiliary be and copula be based upon whether or not be appears with another verb. If it does not appear with another verb, it is a copula, but if it appears in its role to help establish progressive aspect or the passive voice, it is an auxiliary. The sources immediately contradict themselves, however, as soon as they mention the inversion and negation diagnostics for identifying auxiliaries, since these diagnostics clearly suggest that copula be is an auxiliary. My conclusion: the sources are confused. --Tjo3ya (talk) 00:53, 17 September 2012 (UTC)

It seems like the sources might have failed to word their discussions (or just failed to recognize) so that an exception for BE as copula is included. At any rate, I think preempting such similar ambiguity here, that this article could safely provide some clarification by including a section on BE that points it out as a unique exclusion. It's potentially troublesome in two ways, first that people need to be able to recognize when BE is and is not acting as an auxiliary (easy enough, if it precedes another verb in series, it's an auxiliary and if it doesn't, it's not); second, they'd need to be made aware that even when not acting as an auxiliary that BE "passes" the diagnostics. So long as these two things are accounted for, I don't see a problem or any reasonable room to be considered in conflict with the sources as I'm guessing they assumed this would be understood (or at least I hope they did).

Have you taken a look at this Copula_(linguistics)? This article provides a bit more leeway as to what all is grouped under 'copula' and includes statives like seem, become, get and feel. Any discussion of copula within this article might want to always specify copular BE or BE as copula to prevent someone from thinking whatever statement applies to all copulae since the rest of those all have DO as their auxiliary.Drew.ward (talk) 02:29, 17 September 2012 (UTC)

My personal view is that there isn't sufficient empirical reason enough for a distinction between copula be and auxiliary be. The two are one and the same lexical item. There is also confusion, I think, with the terms copula and copular verb. I've always viewed these terms as synonymous, but now that I think about it more, my guess is that linguists subconsciously use the term copula more to denote just be, and copular verb more to denote be as well as verbs like become, seem, taste, feel, etc.

That's all somewhat beside the point, though. As far as this article goes, I am not opposed to expanding this matter into a proper subsection. The short paragraph I just added to the end of the section "Meaning contribution" might be expanded into an entire subsection. That would solidly preempt the possibility of confusion, as you point out. Let's wait and see what Peter has to say. --Tjo3ya (talk) 03:17, 17 September 2012 (UTC)

I feel that I shouldn't have "descended on" this article when I don't have time to make more of a substantive contribution. Note that I only added the tags to show the kinds of issue that exist; do feel free to remove them, of course ideally after adding sources. It does seem to me, though, that both of you (Drew.ward and Tjo3ya) are falling into the trap that I did when I first started editing Wikipedia, namely writing teaching material rather than an encyclopedia article. Linguistics always seems to have been a highly contentious subject; I quite agree that many linguistic texts are inconsistent both with each other and internally. When I taught this subject, it was my job to make choices (without disguising the fact that there were other opinions). But in Wikipedia we must not make choices between mainstream view; we have to explain what the sources say as clearly as possible, giving due weight to different views even when these are contradictory. I'm sure you both know this, so please take this only as a friendly reminder.

On the substantive issues, I think that Radford's definition of AUX is clear that be is only an auxiliary in USE (p. 494). But he doesn't explicitly make this qualification in his definition of copula (p. 501) although he writes that are is a copula in such sentences as They are lazy, implying, but unfortunately not saying, that it would not be in other sentences. This does seem to be a different definition from that in Jurafsky & Martin. I'll look for some more sources, but I think it's uncontentious to say that sources differ in whether the copula be is an auxiliary or not. Peter coxhead (talk) 09:35, 17 September 2012 (UTC)

See below for a comment on some new material I've added, which is intended to start the process of reflecting better the variability of views in the literature of this subject. Peter coxhead (talk) 13:40, 17 September 2012 (UTC)

Let's debate the areas of disagreement

Drew, You are quick to act, too quick, in my view. I provided a response to your original comment from last night, yet you have not responded, but rather you have begun large scale deletions of the content here. I have reinstated the content, and I invite you to first debate the points of disagreement with me before anything else is changed. Perhaps we will both learn from our opposing points of view.

Your difficulty with the term "light verb" is contrary to the literature. You argue that "light verbs" do not exist. Please google "light verb", then read for a few minutes. The term is widely employed, and it denotes a concrete phenomenon.

I can agree to an extent with your point about the use of the term "semantic". There is, however, a broad distinction to be drawn between "functional meaning" and "semantic or lexical meaning". Would you accept the term "lexical meaning" instead of "semantic meaning"? --Tjo3ya (talk) 14:09, 16 July 2012 (UTC)

Tjo3ya, I believe that it's not I who have been too quick to act. You and others are making drastic changes to this article without proposing such ideas in discussion such as these. Your view of auxiliary verbs is very limited and is counter to the syntax of English and many other languages. The generalizations, assumptions, and limitations that you are wording into this article can only be held absent of full understanding of the linguistics of auxiliaries and their roles. Certainly you would not be the first to make such a jump, as the authors of the sources you have cited do exactly that. Saying that my "difficulty" (I assume you mean dispute of rather than difficulty because I would hope you'd not assume I can't understand that) with "light verbs" is contrary to literature is only a partially true statement. The support of this concept only comes from certain sources and does not represent something that is an accepted component of general linguistics nor of English grammar. Rather than making these changes and asserting validity based on obscure writings, why not within these discussions, prove the validity of your beliefs on these subjects? Certainly if they are valid and so widely published about, proving these arguments correct should be easy right?Drew.ward (talk) 15:15, 16 July 2012 (UTC)

Light verbs

Drew,

There is a lot in your response. Let me respond here first just to your dispute with the term light verb. I am currently sitting in an academic library with approximately 20 syntax and grammar books in front of me. The term light verb appears in three of them. In this regard, I think you are more right than wrong when you state that the term is not widespread. The term is, however, backed up by some prominent names, in particular, Otto Jespersen (1965:Volume VI: 117):

Jespersen, Otto. 1965. A Modern English Grammar on Historical Principles, Part VI, Morphology. London: George Allen and Unwin Ltd.

For me, the term light verb is well chosen, and for that reason, I am comfortable with adopting the term into the nomenclature I use and I am in favor of others understanding and using the term as well. A light verb is a verb that is "light" on content. The majority of meaning in a light verb construction is provided by the noun phrase with which the light verb combines. The basic phenomenon is illustrated well across the following variants:

a. She showered. - Full verb

b. She took a shower. - Light verb

a. He joked twice. - Full verb

b. He made two jokes. - Light verb

The similarity in meaning across the a- and b-sentences is concrete for me, and as stated, I think the term "light verb" is good choice to denote the phenomenon, since the verb in the b-sentences is indeed "light" on semantic (or lexical or whatever you want to call it) content. From a truth-conditional semantic point of view, a light verb alone does not build a predicate, but rather it builds a predicate together with another expression, usually a noun. A full verb, in contrast, is alone the core of a predicate. In other words, to have a good understanding of the light verbs and what they are doing, it helps to have a solid understanding of predicate-argument structures.

Since you have a point concerning the extent to which the term "light verb" appears in grammar and syntax textbooks, I can agree that there may not be enough motivation to include the comparison between auxiliary verbs and light verbs in this article. For that reason, I am going to remove the section and adjust the rest of the article accordingly. Concerning the article on light verbs, however, I definitely think that that article is valuable and should stay. There should be a Wikipedia article on light verbs, since the term definitely exists in the syntax literature. When someone comes across the term and does not know what it denotes, having a Wikipedia article that presents the basic phenomenon is very helpful.

As a side note, conducting these exchanges with expletives such as Whoa! and then producing blanket statements such as It's now full of generalisations, assumptions, and references to ideas that are neither accepted nor even remotely established. are not conducive to reaching a compromise. More in my next comment. --Tjo3ya (talk) 18:39, 16 July 2012 (UTC)

• D.J. Allerton in the article in the article "Verbs and their Satellites" on page 176 in Handbook of English linguistics writes the following:

"Choosing a particular structure out of such a possible range clearly affects the grammatical structure of the whole sentence, but what all stretched verb constructions share is the way the process of helping is signaled by an adjective or noun rather than the verb. The verb in such constructions has been variously described as a light or thin verb, or as a support verb, because its semantic contribution is so weak. Verbs can thus range in their semantic contribution from full lexical verbs through light verbs to auxiliary verbs. But the main lexical verb lies at the heart of a sentence and determines its basic network of relationships in that sentence." This suggests that there is nothing esoteric or non-standard in posing a distinction between light verbs and auxiliary verbs in mainstream texts on English syntax.·ʍaunus·snunɐw· 13:57, 14 September 2012 (UTC)

Explicit statement of different views

I've added a short paragraph at the start of Auxiliary_verb#List_of_auxiliary_verbs_in_English which makes explicit the fact that there is widespread disagreement over this category, not just over how to define it and hence what words it includes, but also whether it is a useful category at all. I don't have time at present, but what needs doing now, in my view, is:

• Merging in List of English auxiliary verbs, making much clearer, with references, the variability in lists from different sources.
• Re-writing Auxiliary_verb#Diagnostics_for_identifying_auxiliary_verbs to the same effect, i.e. making clear, with references, the different criteria used by different sources and also why other sources don't use the category at all.

The lead will then need some rewriting. Peter coxhead (talk) 13:38, 17 September 2012 (UTC)

Peter, The notion that there should not be a category AUX is far out for me. The library here where I am has Huddleston (1976). I'm going to see what he has to stay. I will be surprised if he argues outright against the existence of such a category. My suspicion is that he may be arguing against the traditional structural analysis of auxiliaries, where the practice was to assume a flat structure consisting of S, AUX, and VP. I will check and get back on this.

The statements about taggers have injected a significant computational linguistics slant to the initial presentation of the list of auxiliary verbs. In theoretical linguistics and in foreign language pedagogy, the acknowledgement of auxiliaries (aka helping verbs) is completely pervasive. I question whether such statements about taggers should occur up front so soon in the article. A better place to put them might be in a separate section that addresses how auxiliaries are treated in terms of computational taggers and parsers.

Concerning the desire to redo the sections on the diagnostics for identifying auxiliaries, I am not opposed to that. But it looks like we are headed toward a close examination of every possible candidate for auxiliary verb status. I hope you are willing to invest some time in this, because I am. --Tjo3ya (talk) 14:42, 17 September 2012 (UTC

Glad to get you reading Huddlestone, but he's only one of those who have argued that the category AUX isn't useful independent of the category verb. My only point in putting the tagging example early is that taggers represent a modern example of the view that AUX is simply not a useful primary category in English (other languages may be a different matter) and I think this should be said early. AUX is not a category like V or N where there is universal(?) agreement on their validity and not much variation in definition, as is clear from the literature, and this needs to be said from the start, though exactly how and where sourced is another matter. My addition is very much a starting point.

On your second point, not every candidate need be covered in my view; this is a general encyclopaedia not a linguistics one. Just enough to show (a) the main ones (b) that there is commonality but also disagreement. Peter coxhead (talk) 19:46, 17 September 2012 (UTC)

First let me say that I agree with everything both of you have said, and might even quote part of Peter's wisdom on not writing as if teaching always as linguists, which, since we tend to all spend much of our time teaching, certainly is a very apt warning and something I'd imagine it's hard not to do intentionally or not.

Second, I can't help but feel that the potential issues with this article and with the various places that aux is used and that lists of aux are defined, comes more than anything else from a difference in how aux (and for that matter how 'category') is being defined. Subject, object, etc are functional categories; there is no part of speech or lexical entry anywhere that is by nature always a subject or object, but at the same time any range of nouns, pronouns, clauses, etc can function as such. I've never seen much argument for calling on the other.

Verb however is another story. We use the term verb for the part of speech, the functional role, the head of a string of verbs ("verbal catena" Tjo3ya?), that string of verbs in total, and even for everything in an utterance that is "verby". The ambiguity is immediately obvious and problematic with these varied yet related uses because most texts and most arguments seem to never specify whether they are using verb to describe one of the functional roles or as a lexical category, OR both.

Here's my take with AUX: AUX, like subject, object, predicate, etc, is a functional category that refers to the use of an element in a supporting role within a verb construction to convey verbal information (tense, aspect, perfection, voice, polarity, mood, etc) beyond such related information as conveyed lexically by the verb they subordinate ("modify"); further, an AUX may consist of a range of units from lone single-form verbs with no non-auxiliary usage, fully declinable verbs, verb construction (ie be+going), to even phrases (ie be+able, had+better, be+about) but since auxiliaries always fill a verbal position, they must themselves regardless of form, contain at least one verb themselves.

This is how I personally define verbal auxiliaries. I don't think this has been represented in published works other than my own, so I would be leery of putting something like this in myself. But, I do think it accounts for the various discrepancies cited herein and the constant debates over lists of auxiliaries, lists of modals, etc.Drew.ward (talk) 15:28, 17 September 2012 (UTC)

I think that those approaches that use +AUX as a feature which may be present as well as +V (something discussed by Huddlestone for example) are perhaps parallel to if not the same as your definition above, and this has been a very common computational approach. Peter coxhead (talk) 19:46, 17 September 2012 (UTC)

Peter, I've just read Huddleston. He was arguing against AUX as a primary category, not against acknowledging a category of auxiliary verbs. In other words, I think he was arguing that auxiliaries form a subcategory of V. I also think that almost all of modern syntax agrees with him, and I certainly agree with him in this regard. But now I see the problem that the reference to Huddleston that you added misrepresents what he was saying. He was arguing for AUX as a subcategory of V, not against a category for auxiliaries entirely. I suggest removing the removing or correcting the reference in this area. That is important, I think. --Tjo3ya (talk) 20:29, 17 September 2012 (UTC)

Well, that's not how I read him! His positive proposal begins about half way down p. 216. He says explicitly "Our proposal will be, therefore, ... that there is no Aux constituent – that progressive be and may are intransitive verbs taking an embedded complement sentence as underlying subject." On p. 217 he says "We can now dispense with the categories VGrp, Aux and M." On p. 223 "In the new analysis the auxiliary verbs of traditional grammar will in general be simply intransitive verbs taking subject complementation ..." He says (p.218) that the difference between Do you need to tell him? and Need you tell him? is simply a difference in features, not in deep structure. (Whereas in the traditional analysis, need is a (main) verb in the first and an auxiliary (verb) in the second and the parse trees will be quite different.) Huddleston gives need an optional feature which allows it to be fronted, while for can this feature is obligatory and for want it is forbidden. Thus there's no category of auxiliaries; merely a difference (which we know varies by dialect) in whether a verb is lexically marked as allowing movement to the front along with tense. Yes, there's a subcategory of verbs which possess these features (a subcategory which can be marked +AUX in the more traditional approach) but this is merely a superficial categorization: "in the new analysis the I versus II distinction [the distinction between auxiliaries and verbs] is handled purely in terms of rule features" (p. 218). The nature of these features (optional, compulsory, forbidden) differs by dialect; there's no clear-cut distinction between auxiliaries and verbs in Huddleston's approach.

I happen to think that he's right, but that isn't the issue! Peter coxhead (talk) 12:58, 18 September 2012 (UTC)

Peter, I do not have the book here with me and I may not be able to get my hands on it again for a couple of days. From my recollection however, Huddleston establishes a distinction early on between "class 1 verbs" and "class 2 verbs" (I think that was his terminology), as you acknowledge. The distinction rests on the standard inversion and negation diagnostics. Class 1 verbs (traditional auxiliaries) allow inversion and negation, whereas class 2 verbs (e.g. modality verbs like seem, begin) do not. His objection to AUX, according to my understanding, was the fact that Chomsky in Aspects had posited AUX as a primary category in addition to M (class 2). Thus for Chomsky, an AUX (class 1) was not an M (class 2) and an M was not an AUX. Huddleston, however, was saying that AUX and M were subcategories of a single greater category, which is an analysis that most any modern syntax is going to agree with insofar as they can both take a main verb as their complement. In the Chapter summary, Huddleston then comments about the structural status of class 1 and class 2 verbs. He states that they should not be taken as dependents or sisters of V (main verb), but rather if anything, they are heads over V, which is again an analysis that is now widely accepted by most any theory of syntax. At least that is what I understood.

The problem with your addition to the article in this area is that it suggests that Huddleston was saying that there is no such thing as an auxiliary verb altogether, when what he was really saying was merely that there is no such thing as an AUX in the sense of Chomsky (1965). The AUX category of Chomskey (1965) has, I believe, faded, so that your average Wikipedia reader is not going to know what Huddleston was arguing against. My guess is that if we look in Huddleston's more recent works, we will find that he readily acknowledges a category of auxiliary verbs. In fact, that will be a good experiment. I'll check next time I am in the library. --Tjo3ya (talk) 15:43, 18 September 2012 (UTC)

Be careful to distinguish between (1) Huddleston's description of the traditional distinction between what he calls I and II type verbs and his critical analysis of this distinction, and (2) his later proposal which to me is quite clear, as the quotes show. I agree that there are, of course, problems in reading Huddleston now, because his analysis and proposal necessarily relate to the era in which he wrote, which is why I wouldn't use him as the sole source. I can only repeat that the key point which needs to said, and said early, is that unlike categories such as "noun" or "verb", there remain controversies over the category "auxiliary": whether it's a useful category; how it should be defined; what lexemes belong wholly or partially in the category. I'm quite happy for this to be said differently, and supported by different sources as my addition was very much a first draft, but not to say it at all is very misleading. Peter coxhead (talk) 11:45, 19 September 2012 (UTC)

Peter, I've also now checked Radford's definition (1997:494) of auxiliary. He makes the same mistake that he makes in his 2004 book. He contradicts himself by suggesting that there are uses of be that do not allow subject-auxiliary inversion. I have yet to come across a use of be that does not allow subject-auxiliary inversion. All uses of be are auxiliaries. I think we should put that dispute to rest. --Tjo3ya (talk) 20:42, 17 September 2012 (UTC)

You're begging the question by your wording: "I have yet to come across a use of be that does not allow subject-auxiliary inversion." If instead you just say "subject-be" inversion, then you have to decide whether this is a sufficient criterion to call be (in that use) an auxiliary. There may be sources that (a) explicitly say that such inversion is a sufficient criterion (b) demonstrate by examples that they actually mean this – because, as we've noted, the definitions that sources give are not always consistent with the detailed text and the examples they give. However, I don't know any. I certainly don't know a source that says, flatly, "all uses of be are auxiliaries". If there are such, then this should be included, but equally so should contrary views.

Jurafsky & Martin, p. 294, appear to say this, first effectively defining auxiliaries as markers of tense, aspect, negation and mood and then saying that the copula verb be is an auxiliary. But on p. 296 they are clear that have has three "senses": as a main verb, as a modal, and as a non-modal auxiliary. So Jurafsky & Martin, taken literally, are saying that in all uses be is an auxiliary, but have is not. I don't believe they mean this; I think it is just muddled writing on their part. Peter coxhead (talk) 12:58, 18 September 2012 (UTC)

In 2007 I conducted a study in which interviewers asked 2L English learners to put various active voice sentences in passive voice (after they had been made comfortable with how to do this but before they'd been indoctrinated to believe that certain forms "can't be passive". The point was to see what happened when they encountered forms that are supposedly limited to active voice only. Interestingly enough, without prompting, these speakers (adult native German speakers) not only put those forms in question into passive voice but independently created almost identical resultant forms each time. The first data set was followed up by a similar survey of native English-speaking ESL instructors. While the native speakers required explicit goading to "do it anyway" when presented with the supposedly non-passive forms, they still produces remarkably uniform and similar passive versions of these active-only forms.

The two primary examples given were:

John is a doctor.

&

John is sick.

The passive versions returned were generally:

Being a doctor is being done by John. (or, Doctoring is being done by John)

&

Being sick is being done by John.

The reason I mention this is that regardless of slight differences, respondents overwhelmingly intuitively converted the predicate nominative or predicate adjective in constructions with copula BE into activities. Likewise, they converted the syntactically much shorter be into a longer form that was always progressive aspect, with the 'main verb' of the resultant verb string being not be but instead do. Our findings were that this showed that respondents were analyzing copula BE as semantically representing a much more complex idea that the subject be performing the activity "x" (represented by whatever followed be).

If this is true, that copula BE is a syntactically simpler "shorthand" representation of a much longer more semantically complex form, then that could seemingly have an impact on its ability mesh with this AUX+Verb system and its related diagnostics. If especially as our data showed, speakers of modern English tend to think in terms of activities and performance of those activities, and somehow intuit copula be (which seemingly has no aux) to be short for be doing (in which be IS the aux), then it could mean that claiming be is always an auxiliary, and, that copula be therefore passes all of the diagnostic tests for an AUX is valid, justified, and not truly a separate situation from all other aux+verb situations.Drew.ward (talk) 15:08, 18 September 2012 (UTC)

Status of copula be

Peter and Drew, I think we are belaboring the status of copula be too much. We have a source that clearly states that be is an auxiliary, and we have other sources that merely suggest (but do not state clearly) that copula be is not an auxiliary. And as linguists and native speakers of English, we can easily verify that according to the most widely employed criteria for defining auxiliaries (inversion, negation, and VP-ellipsis), copula be is definitely an auxiliary, e.g.

a. Are you hungry?

b. I am not hungry.

c. You are hungry, and I am too.

Since we have a source that states that copula be is an auxiliary, are aware of data like (a-c), and are knowledgeable linguists, it is fully legitimate for us to list copula be as an auxiliary, as long as we also state that one or more sources may disagree and then list a couple of them. This is precisely what the article currently does. Hence I think it accomplishes precisely what it should in this area. Unless one of you thinks that there is a problem with the article in this area and can clearly state what the problem is, we've done our job. Let's move on.

We should move on to the additions that Peter put in yesterday. I think that the statement about Huddleston is inaccurate and should hence be removed or reworded to make it clearer what Huddleston was saying. And the statements about taggers belong in a separate section. I think that information is valuable, but it should not appear where it does. I am in favor of Peter drafting a separate section that provides a bit of overview about how auxiliaries are treated by various taggers and parsers. That would be a valuable contribution in my view. What say you? --Tjo3ya (talk) 16:27, 18 September 2012 (UTC)

I've never really had a problem with it being listed on here. The only thing I'd recommend including would be a few lines that explain why be tends to get included (because of the way it passes those tests) even when not filling that functional auxiliary role (aka even though it doesn't appear in conjunction with a non-auxiliary verb like all the others do). That "hey wait a minute" observation is the only potential problem I could foresee a reader having with the article (and sources) including both auxiliary use of be (as in vocal "be given", aspectual "be writing", or modal "be to see") and non-auxiliary (as in not functioning as an auxiliary in conjunction with a further "main verb") use (copula). As long as that's justified somewhere in the article I don't see where anyone could have a problem with it.

There is one other thing that I'd like to see included probably in the intro that I think is a bigger issue than the higher-level debate over how to treat copula be. That is that there be something that points out the two main uses of auxiliary -- the more narrowly defined classification that we've been discussing thus far, and also the functional use of the term to refer to the range of linguistic units that can fit the "auxiliary" position to convey whatever their presence supports (mood, aspect, voice, whatever). This second usage allows the article to account for things "be+going", "had+better", etc. Especially with modals, there are far more auxiliaries that have internal structures like these versus those few Germanic modals that would also meet the definition used in the rest of the article. I don't think this article should try to fully include and discuss the range of these forms though because they introduce a variety of levels of syntax and such. For instance, both "had+better" and "be+going" have "standard" auxiliary verbs within their structures. And, those verbs would still function like their other counterparts in regard to the diagnostics. But, unlike a single-part auxiliary where the whole thing moves, with these only the internal aux moves. That gives you an aux within an aux situation where a single unit within the whole construction takes over the various roles normally assigned to the left-most aux in a verb string, but with the "whole aux" in which it appears filling a single auxiliary position overall. It's not too bad with "had+better" which has only a single form that the aux can take. But, when you get to "be+going", aux be can take on any form that be could take in any other situation in which it is acting as the progressive aspect aux (ie 'am/is/are going', 'was/were going', 'will/should/must be going', 'have/had/could have been going', etc). Too much for at least this article to bother with, but since many discussions of grammar (at least English grammar) deal with the function of these multi-unit constructions as filling various auxiliary roles, it seems that some discussion of this functional auxiliary use of auxiliary should be accounted for in the article. It could also prevent confusion or disputes about the main topic of the article from readers who may arrive at it looking for this other use.Drew.ward (talk) 17:43, 18 September 2012 (UTC)

Drew, the second sense of auxiliary verb that you mention does exist, I believe. I think it is probably used more by language instructors than by formal linguists. It goes without saying that if we want to include this second, much broader notion of auxiliary verb into the article, WE NEED good SOURCES. I am open to establishing a "narrow" definition of auxiliary verb, which is what the article currently does, and a broad definition of auxiliary verb, which is what you seem to be suggesting and which relies on backing it up with some good sources. In this regard, I think finding good sources that discuss be going and the like as auxiliary verbs/constructions will not be easy, which means I may resist expanding the article in that direction.

For me, a form like am going consists of two verbs, the auxiliary am and the light verb going. The two together form a catena and are hence a concrete unit of syntax and can hence be construed together as contributing a concrete piece of functional meaning (e.g. near future). I would never say that am going is one verb. I might call it an auxiliary construction or an auxiliary catena, though.

Similar statements hold for had better. The verb had is an auxiliary, and the two words together form a catena, which means they are again a concrete unit of syntax. I would not call this catena an auxiliary verb, though, because it consists of more than one word. Interestingly, the verb had in had better definitely has auxiliary status, whereas if better is absent, have alone (i.e. without a participle) does not usually behave like an auxiliary, although Peter might disagree.

I know there is a significant contingent of language specialists who will employ the word verb to denote multi-word combinations. I personally frown on this practice, since the term verb should denote a single word. We see this difficulty with the term phrasal verb, for instance. A phrasal verb is usually neither a phrase nor a single verb, but rather it is a combination of two or three words that usually cannot be construed as a phrase. These words do, however, form a catena. --Tjo3ya (talk) 19:12, 18 September 2012 (UTC)

Multiple Auxiliaries

I have a couple of questions/comments about this bit:

"The particle to is included in the verb catena, since it often obligatorily introduces an infinitive. The hierarchy of functional categories is always the same. The verbs expressing modality appear immediately above the verbs expressing aspect, and the verbs expressing aspect appear immediately above the verbs expressing voice."

Comment - the copy editor in me just doesn't like the way "since it often obligatorily introduces an infinitive" sounds. I also think (although it's perfectly correct) that you may lose some readers at obligatorily. How about something that is grammatically more simplistic like "because its use is often required with certain infinitives" or "because it is often featured as a required element of certain infinitives"?

Question - regarding the same line, keep in mind that with some auxiliaries there is a difference in function or meaning (usually different moods being expressed) if to is present versus not. For a couple it's seemingly optional with no semantic change such but with its presence having syntactic implications (consider dare and need where not including to results in dare/need filling the aux position (sub/v inversion, negation, etc) whereas adding to while having the same meaning converts dare/need over to a full-verb auxiliary where do now fills that inversion/negation slot (but a form in which the full gamut of TAM is possible whereas the non-to usage is constrained to unperfected present tense, simple aspect. Then you also have those odd ones like ought which are always single-form defective auxiliaries, yet can be used with or without to. Unlike with dare/need where the presence/absence of to dictates syntax, with ought it is the chosen syntax that dictates the presence/absence of to ("You ought to go", "You ought not go", "Ought you not go?", "Ought you go?"). Just wondering if this is accounted for in DG or in the sources?

I am not aware of any sources in DG that address these idiosyncratic aspects of need, dare, and ought, but my guess is that such idiosyncracies are indeed addressed by the computational linguists who are employing dependency-based systems for various applications. We'd have to contact them to get a definitive answer (which might be possible if its important). --Tjo3ya (talk) 18:16, 19 September 2012 (UTC)

Final Question - In "the verbs expressing aspect appear immediately above the verbs expressing voice", the order is correct, but the chart tends to treat the past participle as the 'verb expressing voice'. Yet, in English, the vocal auxiliary be (which requires its subordinate be in past participle form) is the primary marker of passive voice. And, that auxiliary be has to be marked for agreement, tense, aspect, mood, perfection & polarity so that it matches the same attributes of the main verb (now the subordinate of vocal auxiliary be) or verb catena for which that verb is the head in every way. Shouldn't be then be treated as the primary "source" of 'passive' rather than the main verb?

No. The passive voice is very often expressed without the presence of be, for instance when the participle modifies a noun: the revised paper, the things said last night, etc. Further, many languages lack the equivalent of be. For these reasons, I think it is appropriate to view the participle as more essential to expressing the passive than the auxiliary be.

I would have analyzed those examples differently with revised and said last night as adjectival.Drew.ward (talk) 19:56, 19 September 2012 (UTC)

Your statement about the "obligatory" marking of the auxiliary be of the passive also seems inaccurate. As in the example trees in the article, the auxiliary be of the passive appears without any markings for agreement, tense, aspect, mood, perfection, and polarity. It largely lacks these markings every time it appears as the infinitive be. These grammatical categories tend to reside higher up in the verb catena, most of them being expressed primarily by the finite verb. --Tjo3ya (talk) 18:16, 19 September 2012 (UTC)

I see now I left part of my original statement out which should read "And, that auxiliary be has to be marked for agreement, tense, aspect, mood, perfection & polarity so that it (vocal auxiliary be) matches those same attributes of the main verb (now the subordinate of vocal auxiliary be) or verb catena for which that verb is the head in every way, just as it would be marked in active voice." Within this view, if the active voice construction is "may have been painting" then the passive voice equivalent would be "may have been being painted" with everything in italics collectively acting as the vocal auxiliary but with that auxiliary declined to match what the main verb would have been declined as in the active voice version. This is actually the instructions given in many grammar guides for "how to form the passive".Drew.ward (talk) 19:56, 19 September 2012 (UTC)

Side Note - If your description of syntactic hierarchy as modality->aspect->voice is standard throughout DG, it would seem to lend credence to my analysis of multi-part modals as "may be being judged" in your example while having its own internal hierarchy that meets that specified order, actually fills a single modality slot in the hierarchy terminating with deceived.Drew.ward (talk) 07:54, 19 September 2012 (UTC)

Drew, The hierarchy modality > aspect > voice is accepted by morphologists and syntacticians in general, I think; it is not unique to DG. I can produce a prominent source or two backing up that claim, if you are interested.

But I do not fully understand this comment. I think the confusion may be due to the fact that the constituency-based concept of syntactic structure (which is still dominate in the study of syntax and grammar in general) is subconsciously influencing how you think about these grammatical categories. The term "multi-part modal" is an oxymoron for me. A modal is a single word (may, might, should, will, etc.); it cannot have multiple parts in syntax. The difficulty here is that traditional constituency-based syntax has never been in a position to view many catenae as constituents, since they often clearly fail to qualify as constituents. What this means is that for constituent-based theories of syntax, there is no good way to talk about verb catenae such as may be being judged. Hence the terminology is all over the place (How you use the terminology, Drew, is a good example of what I mean). One tends to denote many a multi-word catena as "a verb" (e.g. had better, am going) or "a modal" (e.g. may be being judged), since one cannot accurately call these combinations constituents. The catena unit overcomes these problems, since it is a concrete unit of syntactic analysis. These word combinations are catenae. --Tjo3ya (talk) 18:16, 19 September 2012 (UTC)

Perhaps as the catena takes hold throughout linguistic fields, it might provide a way to rectify such treatment of multi-part constructions that serve a single-unit role. As to the point of the comment, I meant not to call "may be being judged" as a whole a single modal (I think it is, but wasn't trying to push that point), but rather that that entire chain fills the modality position in within the hierarchy of the verb string that follows it. Consider the original example with some substitutions:

"Fred may be being judged to have been deceived by the explanation."

"Fred could have been deceived by the explanation."

"Fred is thought to have been deceived by the explanation."

In all of these examples, my analysis would be that the boldfaced part is expressing a single mood onto have which is perfecting been which is conveying both passive voice and simple aspect with deceived being the content verb for that entire verb construction. This (that 'slot' conveying a mood) fits with your stated hierarchy. For me, it doesn't matter how complex the form that that modal takes, it still conveys a single mood onto the element it subordinates. The "root" modal would be judge which maintains its full verbal flexibility in conveying that mood. It's use in the example has it in a form in which may conveys a potential/possible mood onto the vocal auxiliary be being which is itself put in progressive aspect, and finally the original root modal in past participle form due to it following the vocal auxiliary as judged. As you have done in your diagram, you could easily show this same hierarchy of mood-aspect-voice within the structure of this modal, but that particular hierarchy is internal to the modal with the entire "may be being judged" acting as the modal in the primary verb string "modal have been deceived".

Yes, the catena unit can accommodate what you are suggesting here. But what you are suggesting is very unconventional, and I doubt that a coherent source could be produced to support your preferred analysis. Above all, what you are calling "mood" is what most would call modality. Hence a verb catena such as may be being judged to expresses modality, aspect, and voice. It does not express mood in any way. The only mood that is really present is the indicative. --Tjo3ya (talk) 22:31, 19 September 2012 (UTC)

That's what I am talking about when I refer to a multi-part modal filling a single modal slot. May I ask, within your view: "A modal is a single word (may, might, should, will, etc.); it cannot have multiple parts in syntax.", how would you deal with periphrastic equivalents to single word modals: "can = be able", "must = (do) have", "will = be going" "may (permissive) = be allowed", etc.? These forms developed alongside the rise of the single-word forms and were logical innovations to take over for lost versatility as the traditional modals lost their ability to convey the TAM range of full verbs. Note that in German, these same modals exist, but with those modals maintaining their full versatility, there is no need for and thus no similar system of equivalent periphrastic forms. If you only count can as a modal, how would you employ that modal to express that modality in the future or perfected, or as contingent on something else? You can't, and that's why English developed "be able", which as "will be able", "has been able", "would be able", or "would have been able" conveys the exact same mood but permits the versatility lost in reduction of the original modal to its limited can/could form.Drew.ward (talk) 19:56, 19 September 2012 (UTC)

I don't think I necessarily disagree with the main point here, if I understand it. The periphrastic forms you are considering are of course all catenae. See the trees in the article periphrasis. I agree that can and be able are basically synonymous in the modality that they express. Note that "modality" is something different from "modal": can is but the be in be able is not a modal, whereas both express the same modality. You seem to be interchanging the terms modality and mood consistently. I may be able to agree with your use of the term modality, but your use of the term mood is a bit of a mystery for me. Drew, I cannot go over these things any more at present. I'm spending too much time on this. Let's slow down with this exchange. I will probably not respond here for a few days. --Tjo3ya (talk) 22:31, 19 September 2012 (UTC)