Sunday, September 20, 2015


There is a lot of good in prestructural linguistics that is coming back to the front


This linguist is a descriptive linguist working on British English, though he is not British. This book is entirely based on one book by novelist Sybille Bedford on the longest murder trial in England at the time at the Old Bailey in March 1957. The title of the book was The Best We Can Do in the UK and The Trial of Dr Adams in the US. This gives absolute unity to the corpus used here, and yet that corpus is very circumstantial since many people of different cultural circles met and spoke in this court, some of them having to use a very standardized legal language or even interrogation language: you do not speak in a court of justice the way you speak in the street. And the whole thing is seen, through the pen and the mind of one person who is a novelist, hence a creative writer who has her style and cannot depart from it. Some of the remarks I will do depend on this fact. Martin Joos would have been well inspired to widen his corpus to other British genres, not to mention of course American English. The circulation between the two dialects of English started to be very important in the late 1950s thanks to tourism and to television and cinema exchanges.

My first remark will be on what he calls the “gerund” that he opposes to the “present participle.” He reduces the present participle to a simple adjective and eventually an adverb. He is more or less scrambling his mind when this present participle has an obvious subject and wonders whether it should be the subject personal pronouns or the object personal pronouns. The case of these pronouns is determined by the function they hold as to the main verb in the main clause and the present participle cannot be a gerund because it is a simple apposition to that pronoun, or a noun, hence an adjectival modifier. When I say:

“I don’t like Peter/him reading Hamlet.”

The case of the pronoun (or the noun) depends on their function as to the verb “like” and in no way as to the present participle “reading.” In fact I don’t like the person known as Peter, circumstantially engaged in the reading of Hamlet. This explanation introduces the third use of the Ving form, but more in a moment. We have a present participle when the “subject” is NOT in the genitive and when the object is in the normal objective case using no preposition to introduce itself because it is a direct object, the way it is in my example.

That makes Martin Joos slightly fuzzy when he deals with the gerund: he accepts the two cases as for the subject, both objective case and genitive. He does not see that in the first case we have a real present participle that is apposed to a noun or pronoun giving him thus a deep subject function it does not really have in the sentence. This apposition behaves like an adjective but it keeps its verbal syntax as for its own complements.

The real gerund has a real subject in the genitive case. The genitive is the case of the agent in the nominal field whereas the nominative is the case of the agent (subject) in the verbal field. But this gerund keeps its verbal syntax as for its complements. It can even take adverbs:

“I don’t like Peter’s/his dramatically reading Hamlet in every meeting of ours.”

These first two cases are clear: they deal with the action of reading itself and nothing else.

The third case is not at all considered and Martin Joos misses a point then. The third Ving form can be a real verbal NOUN, a NOUN built on a verbal root. But then this noun has a nominal syntax both before and after. It can take genitives and adjectives before or simple articles, and only nominal complements after introduced by prepositions.

“I don’t like Peter’s amorous reading of Hamlet as if it were a romantic if not melodramatic story.”

Obviously we have shifted from the action of reading to the interpretation (as an actor or as a critic) of the play. We can note the play itself is introduced with the preposition “of” because it no longer is a direct object but a nominal complement to a noun.

My second remark is going to be extremely positive. The way he spreads out the structure of the verbal phrase (reduced of course to the verb and its direct auxiliaries and not including anything else like complements or adverbs). Page 55 he exemplifies this structure in a clear table and page 76 then page 81 he summarizes the whole approach in two clear tables. He thus considers the verbal phrase can be decomposed in six successive verbal elements from left to right (seven if we consider the subject):

[0- Subject, my addition]

1- “Tense” that can be actual and remote, opposing the unmarked present to the marked preterit (-D).

2- “Assertion” that can be factual or relative, opposing simple tenses or the use of modals of various types (WILL, etc.).

3- “Phase” that can be current or perfect opposing simple perspective tenses looking at the action as a whole and the perfective phase as he calls it which looks at an action that started in some past moment and is still going on at the moment of utterance (HAVE-N).

4- “Aspect” that opposes generic to temporary, meaning the full action as opposed to an action seen in its temporal development and as unfinished at the time of utterance or reference. We are dealing here with the progressive form (BE-ING).

5- “Voice” that opposes neutral to passive and the fact that he does not say active is a good thing because there are quite a few verbs who do not imply an action with an agent (“He lies on the ground” as apposed to “He lies to his father all the time.”) or some intransitive verbs who cannot be set in the passive because they are both active and passive (“He runs everyday at seven a.m.” “He” is obviously both the person who is doing the action and the person who is being transported by this running person from point A to point B. To introduce a passive would be tricky: “He is run by his brother every morning for thirty minutes, doctor’s orders!” on the model of “Paul is walking the dog” or “The dog is walked by Paul every night.”). We are thus dealing here with the passive voice (BE-N).

6- “Function” that opposes the propredicate (mostly DO or any anaphoric verbal reduction for simple back references to previously used verbal phrases and actions (“Paul cannot really have been working all night since he looks fresh and rested, can he? – Oh yes he has!”), hence to a full verb (V).

What his presentation does not make visual is the fact that at every level what he gives as the second element is in fact governed by the first but on the next level, as shown in the following table.


à Ø or –s
à -D

WILL, etc.

à Ø

à + Ø
à + TO

Perfect: HAVE

à Ø
à -N

Temporary: BE

à Ø
à -ING

Passive: BE

à Ø
à -N


I have kept most of Joos’s terms though I would favor others. He then exemplifies his approach by using 0 for the first choice at each level and 1 for the second choice at each level. The most complete example he gives is 111101: “She (1) would (2+3) not have (4) been (5) having (7) much.” He does not provide an example for 111111 and that is probably because his corpus does not contain one that could be in the same genre: “She (1) would (2+3) not have (4) been (5) being (6) murdered (7) when I arrived if the murderer had been arrested after his first crime.” Note the shortest possibility is “Paul (1) runs (2+7).” Due to the neutralization of 3 to 6 strata.

The third remark is that he does not approach modals properly. He does not approach them syntactically but only semantically and thus he cannot understand the opposition between CAN in “Yesterday he could run two miles” with the alternative “yesterday he was able to run two miles,” and MAY in “Yesterday he may have run two miles: I don’t actually know” with the alternative ‘Yesterday he might have run two miles: I don’t actually know.”

The first use is deictic because it refers to a real action that was possible and was performed. In the second case we are dealing with an epistemic modal since the prediction of a probability is always positioned in the present (a neutral probability with “may” and a reduced probability with “might,” both reflecting the level of prediction implied by the utterer) and the action is positioned in the past by the perfective auxiliary “have.” We are dealing here with syntax and not semantics and it is these syntactic structures that determine our understanding of the modals. The chart presented above then has to be modified to integrate this case.

Actually Martin Joos, due to his corpus I guess, practically only examines the permissive value of “may” and “might” though he calls it the “archaic sense.” But he does not seem to really explore the non-archaic sense, the sense of probability. It’s obvious the permissive  meaning is archaic but what’s more it is absurd in sentences like: “The doctor said the baby may die tonight.” It cannot be a permission, of course not. “Might” will make the probability more relative, hence lower or higher but a permission is out of place in the context and the use of “may” and “might” for probability prediction is absolutely modern and common.. Same thing of course in sentences like: “The parents were extremely anxious about their baby who may die tonight.” No one has given the permission that the baby should die tonight which would become an order and hence a criminal action.

The last remark is about “shall.” Here again he is conscious that “shall” is not a simple future auxiliary. He is conscious that it is not a simple first person auxiliary. He is also aware that “shall” in an assertive sentence is not the same thing as in a question or in a suggestion (“What shall it be, gin or vodka?” and “Shall I suggest a glass of Port?”)

But he is entangled in court language in Great Britain in the 1950s when “shall” was still used as a future auxiliary, marginal, for all persons, but yet still a future auxiliary. Since then this value has practically disappeared but contractual language has valorized the basic meaning of “shall” as being the future prediction of something that depends on some authority of some kind over the person who is supposed to perform the action, which gives in the interrogative form the value we have just quoted (someone asking a person what he/she would like to have or a suggestion from someone but submitted to the agreement of the other person.) It is the basic value and has always been. In King James’ Bible we have: “Thou shalt not kill.” (Exodus, 20:13) and it is systematically used in second person singular or generic (plural) in that version of the Bible. Today it is only used in situation where someone is confronted to a strong authority that dictates their actions. “Of course he shall do it, since I am his boss on this question.” And there is no other way to say it with that meaning and that force. In a contract it has exactly the same value that “must,” “have to,” or “ought to” do not carry. To have the same power as “shall” we should say something like: “you must absolutely do this.” You do not find such a phrase in a contract or even in a user’s manual of some industrial machine. In fact in a user’s manual the equivalent could be a simple imperative.

He is right though that “shall” should be used in contexts that absolutely require that modal, and these contexts are rare and certainly not in colloquial everyday language. The only common use is in the interrogative form though it is always a little bit stilted and formal. In the underground if someone is blocking the way out of the carriage you may say: “Shall I push you or shall you move?” but you might be confronted to some surprising reactions. Not everyone speaks like the upstairs gangs of Downtown Abbey.

Apart from that it is an interesting and useful book, even if it is not a highly computerized model of the English language. At times it is necessary to wean ourselves from all Universal Grammar of any type that turns us into some translating machine producing gibberish codes of signals instead of language.


Comments: Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?