" The Architecture of English Grammar : from the Theory of Phases to the Principle of Cyclicity"

Conférence plénière (en anglais) au colloque de linguistique anglaise à Toulouse, Juillet 2000. Ce texte fait le point sur la genèse et de le développement de l'approche métaopérationnelle.



Introduction :

            The idea that language is a system is not a new idea. Systematicity characterised the studies of the Indo-European family and Grimm’s Law is generally the students’ first encounter of the fact that languages are organised wholes. The principle of exceptionlessness (Ausnahmslosigkeit) put forward by the Neugrammatiker is another example of the systematic behaviour of our tongues. Humboldt’s " innere Sprachform " transfers the systematicity to the very essence of languages, their internal structure. For the author of " Uber die Verschiedenheit des menschlichen Sprachbaues " a language cannot be t a u g h t : what one can do is " den Faden hingeben " that is " give out the thread " and the system will unroll by itself.

Perhaps the most often quoted definition of language was coined by Antoine Meillet : " La langue est un système où tout se tient et a un plan d’une merveilleuse rigueur " (A language is a system of interdependent parts which displays a plan of a marvellous rigor). The word " system " is certainly the most important word in Saussure’s " Cours de Linguistique Générale " : "la  langue est un système de signes… ", "  un système basé sur l’opposition des impressions acoustiques " etc. The structuralist school inherited what had so far remained an interesting intuition and tried to achieve in grammar what Roman Jakobson had brilliantly done in phonology. Structuralist grammarians began a furious quest of the basic syntactic structures , the finite set of patterns out of which our utterances were built up (cf. Paul Roberts’s series of grammars or" Les Structures Fondamentales  du Français", Bernard Pottier , Nancy 1962). But the best achievement of that period was undoubtedly Martin Joos's "Cube of the English Modals", an original systematic construction. From 1957 onwards generative grammar has been but another attempt at capturing the web and woof of languages. Let me recall the short-lived notion of "kernel-sentences" , which, in the first chomskian model, were considered to be the simplest possible declarative active sentences. Generative grammarians transformed the linguistic landscape by introducing abstract tools and by breaking up the linear conception of utterances of their structuralist predecessors.

But as far as I know the only linguist that tackled the concept of system seriously was Gustave Guillaume. His immense work - fifteen volumes of his "Leçons de Linguistique" have been published so far - was a powerful attempt at demonstrating the systematic organisation of languages . He was certainly the first linguist to oppose the facts of discourse, the visible facts, to " a network of secret invisible facts, much more important than the surface observable facts for the comprehension of the essence of a language". His 1952 paper "La Langue est-elle un Système?" is certainly one of the highlights of linguistic literature.

Meanwhile descriptive grammar ruled unperturbed over the grammatical field , although it was nothing but a mere  juxtaposition of independent unrelated problems . Zandvoort's "Grammar of Contemporary English" opened with the traditional chapter on the different sorts of infinitives (plain or full) and linearity triumphed throughout the book : the chapter on the "defectives" is an excellent illustration of the Dutch grammarian’s linear approach. But the most important achievements of descriptive grammar were the two monumental  grammars  written under the direction of Randolph Quirk and published respectively in 1972 and 1986 : “A Grammar of Contemporary English” and “A Comprehensive Grammar of the English Language” . These two treatises were certainly the best and most complete descriptive grammars of English but at the same time they clearly showed the limits of the descriptive approach. French grammars also remained well behind the linguistic ideas of the time: Marc Wilmet's "Grammaire Critique du Français" (1998) - to say nothing of Chervel's "Histoire de la Grammaire Scolaire du Français"(1977) -gives a provocative account of the type of grammar that was being taught in secondary schools and at the universities.


My personal linguistic adventure started  at the beginning of the seventies when I decided to choose the BE+-ING construction as the subject matter of my thesis (thèse d’état). The reason of my choice was my dissatisfaction with the current treatment of the so-called « progressive form ». I simply could not teach the traditional viewpoint to my students ! I felt there was more to it than met the eye and that the problem was far from being solved (in fact it looked as though it had been definitely shelved ! ). My conviction was strengthened , quite unexpectedly, by the writings of Martin Joos, first « The English Verb » and then his review  of  A. Ota in LANGUAGE (vol.10, 1964) where he called the traditional explanation (« action making headway ») « preposterous  » !

My aim at the time was far from revolutionary . What I wanted to find was a better treatment than the one proposed in grammar-books, that is  a fundamental m e a n i n g  that was compatible with the varied occurrences of the form (although I appreciated Joos’s criticism of the weaknesses of the « progressive » or « continuous » form, I found that his solution did not differ substantially from the current one : his « limited duration of the predication » was not in my opinion the key to the problem).

In 1973  I presented the results of my three-year research at a pedagogical conference to an audience of two hundred inspectors and experienced teachers of English. Confronted with the diabolical minimal  pair :

      (1)   I leave tomorrow

(2)      I am leaving tomorrow

 I dared  speak of the formal role of –ING, a metaoperator  signalling a sui generis type of syntactic linkage, instead of the anecdotal pseudo-analysis usually put forward, i. e “.You say (1) when you talk to your secretary  and you use (2) when you want to apologize ».

My analysis started with  a redefinition of the SCOPE of –ING : -ING  did not bear on the VERB ALONE but on the WHOLE VERBAL GROUP, so that in (2) we had to deal with the complex verb LEAVE TOMORROW (-ING (leave-tomorrow )). So, (1) was to be analysed into three components ( I - LEAVE – TOMORROW) whereas (2)  resulted from A BINARY RELATION : I / LEAVING TOMORROW, linked up by the operator BE.

The traditional parallel analysis of (1) and (2) is unfortunately a « trompe l’œil », the result of what can be called the linear fallacy : 

      (1) I                    leave                    tomorrow

      (2) I                    am leaving           tomorrow 

What I found out was that in (2) the status of TOMORROW was quite different from what it was in (1). : in (1) TOMORROW belonged to an OPEN paradigm ,  that is it could have been TONIGHT or ON TUESDAY or IN TEN DAYS’TIME whereas in (2) TOMORROW represented a CLOSED paradigm and was part and parcel of the complex form LEAVE TOMORROW. So, the analysis of the pair (1-2) should be read like this : 

      (1)  I                          leave                tomorrow

      (2)  I                          am                   leaving tomorrow

 My analysis corresponds in fact to the intuitive meaning of our minimal pair : a simple announcement of departure in (1) and  the expression of a  an antecedent decision which may  serve as an excuse for not accepting, say, an invitation to dinner . In (1) what matters is the date of the departure  but in (2) the utterer - speaker is speaking of himself, since the –ING verbal group is orientated towards the grammatical subject « I ». The role of –ING is to show the THEMATIC status of the verbal group, a condition of the application of that group to the subject on  the left.

The linguistic consequence of these preliminary remarks is that the grammar of English opposes two different sorts of predicational linking : either  a primary rhematic type of utterance or a secondary thematic one where the metaoperator –ING blocks any paradigmatic choice in the verbal group. This is how the theory of the two phases was born.The thematic kind of predication is triggered by such factors as situation, anaphora or simply syntactic cohesion. These presupposing  conditions  are responsible for the emergence of the –ING verbal group. We shall see later on when speaking of the V1V2 problem that there exist presupposing verbs, such as KEEP, AVOID or RISK for example which demand a PHASE 2 verbal group (V2-ING) : He kept looking at the clock ; you must avoid drinking alcohol when  taking antibiotics ; he risks losing his job.

 It should be borne in mind that this theory wants to reflect the processing of the utterances, for it is at that level that the opposition OPEN paradigm / CLOSED paradigm applies.

Let us examine some utterances to let this viewpoint sink in :

(3) * I left tomorrow.

(4) I was leaving tomorrow, but now I won’t. 

The agrammaticality of (3) and the bona fide (4)  are clearly the result of the different status of TOMORROW in those utterances.

 (5) If that man comes into my home, I leave it ! 

Here is a good example of the use of Phase 1 : LEAVE has not been announced  in any way and is the choice of the moment. *I am leaving it would be agrammatical here.

The same is true of (6) below : 

(6) I suggest that we leave tomorrow.  

On the other hand BE+ING is obligatory in (7) and (8) : 

 (7) He returned carrying a suitcase. « You are leaving ? », I asked in some surprise.

(8) I’ve got a bad news for you : the charwoman is leaving

A brief incursion into  French  will show us the ambiguity of an utterance like (9) : 

(9)   Je pars demain. 

This sentence yields two possible analyses :

(9’)            je                        pars                demain        (I leave tomorrow)

(9’’)           je                        pars demain                       (I am leaving tomorrow)

As is well-known this type of ambiguity concerns the present tense only, for the distinction passé simple/imparfait will make up for the  absence of a BE+ING construction in French grammar.

The Theory of the Two Phases will be developed in my thesis ( « BE+ING dans la Grammaire de l’Anglais Contemporain »1976) and in the grammars of English that I published between 1982 and 1996. Here is the schema of what was to become the starting-point of the Double Keyboard : 

            PHASE 1 : rhematic (open paradigm) 

      PHASE 2 : thematic (closed paradigm) 

Note : My use of RHEMATIC and THEMATIC has nothing in common with the traditional theme-rheme opposition proposed in 1929 by Mathesius and which has been made great use of  since then (in particular by Halliday in the sense given/new). My phases do not refer to the extralinguistic world but to the internal  organization of grammar .

Here are some important points to be noted to make things clear :

1-     My analysis of the BE+ING constructions (with –ING as an INVARIANT formal operator) shows the pernicious character of the commonly used label S V O (Subject- Verb-Object) : SVO is a product of what I called the linear fallacy, since it cannot capture the difference between the minimal pair (1-2) and the difference between ternary and binary predication . The same criticism applies to Chomsky’s type of derivation  at the beginning of the sixties : how could one signal the different status of TOMORROW in (1) and (2) on  a syntactic tree ?

2-     The  purely formal character of the invariant tracer –ING must be strongly stressed :        -ING belongs to what I consider to be the inner metalanguage of languages. Its contribution to meaning is indirect (but remember I started my research with the aim of finding a basic MEANING of the BE+ING form : I was a fortunate victim of serendipity!). The theory of the two phases enabled me to have a fresh look at the so-called simple forms of the verb and in particular the « simple present » (the PHASE 1 PRESENT). Here are some examples contrasting minus and plus BE+ING utterances : 

(10) - Is anything the matter ?

- Pardon ?

-         I am asking you if there is anything the matter

    Anaphora is a frequent trigger (far from automatic !) of the BE+ING construction. Sometimes the grammatical mine needs no digging whatever : this is the case in utterances of the following type :

 (11)When you fly the Concord you are flying in the fastest plane in the world.

(12)a) I protest , Mr.Chairman, this is unfair.

(b) Railwaymen are on strike. They are protesting against the new regulations.

(13)a) (A religious procession with the statue of a saint). Suddenly, somebody shouted : « She weeps ! She weeps ! »

b) He raised his eyes and looked at me. He was weeping.

The backgrounded character of BE+ING utterances highlights the  fact that simple present forms lack the particular filtering of the utterer, which  explains the agentive role of the grammatical subject in the rhematic phase. I remember the embarrassment of many teachers of English when confronted with a simple present which had nothing to do with the habitual  meaning proposed by  their grammars :

 (14) And now the Queen walks to the throne (  an utterance heard during the Coronation of Elizabeth II, 1953 , which provoked a row  among teachers of English in France , for it did not  match with their conception of the “simple present”!)

Even the simplest utterances of English cannot be accounted for by the current grammars of English , sentences such as  the following , for example :

(14’) Here comes the bus or :

(14’’) Belmondo cuts his birthday cake (legend of a photo representing the scene).


Let me add that this particular weakness of practically all the grammars of English available in 2000, has dramatic consequences on the didactics of English. Even the latest manuals for first-year learners of English still stick to the traditional but false opposition “habitual versus real present”(simple present vs. continuous present).

 3-     I cannot enlarge on the total inadequacy of the traditional pseudo-explanations of the progressive (continuous) form, but let me give one example of the counterfactual character of the « imperfective » theory (imported into English from Russian grammar) :

 (13) We saw Mrs. Lucas on the way. She was carrying big bags of shopping. My mother waved but Mrs.Lucas couldn’t wave back . (Adrian Mole’s Diary)

The « preposterous » character of the  « unfinished activity » explanation needn’t be underlined.What the author (Townsend SUE) meant was the predicament which prevented Mrs.Lucas from waving back !

 -4 – Utterances of the IS BEING type present no difficulty in the analysis I am proposing. The open/ closed paradigm theory accounts nicely for pairs like the following : 

(15) a) Where does he come from ? – He is French.

b) He isn’t being nasty. He is just being French. (The Time Magazine about the former French Foreign Minister, M. Jobert).

The status of the adjective FRENCH is quite different in a) and b): in a) it is being used rhematically (open paradigm of possibilities) whereas in b)  FRENCH is being used thematically (no paradigmatic choice).

 5- The verb WEAR offers an excellent exemple of the importance of the utterance orientation : towards the object (PHASE 1) or towards the subject (PHASE 2) :

(16) He wore a carnation at night. He always did.                                                             (16’) She had been at church and was wearing a hat . 

(16’’)- How are you to recognize Mr. Montgomery ?                                                              - He will be wearing  a white makintosh and a black hat (J.H.Chase).

 6- BE+ING utterances containing such adverbs as : always, perpetually, ceaselessly etc.  have always  baffled traditional grammarians  . Even Martin Joos had to admit that « She is always grumbling »  was « a teaser » if you considered « limited duration » to be the key to BE+ING constructions.That kind of utterances  present no problem whatever in my type of treatment. In the following examples we find the typical binary relation underlying all BE+ING utterances . As to the modal adverb its scope is the linking verb BE :

 (17) – Mrs.Smith has seen a ghost again.

-         Oh, she is always seeing ghosts !

      (18) He is perpetually quarrelling with his wife !

      (19) This girl is always knowing something she isn’t  supposed to ! (an utterance which embarrassed the Belgian researcher E. Buyssens in his thesis : “Les deux aspectifs de la conjugaison anglaise”).

 Traditional grammarians  attributed the BE+ING form to « irritation »- another example of what I an calling “direct assignment of meaning”. Well the irritation is certainly there but it  is  the result of the modal adverb multiplying ad nauseam the grammatical link BE (ALWAYS  can be a frequency adverb (I always smoke a cigar after a good dinner) or a modal adverb as in the utterances above ! This is  a  good example  of  the power of an authentic grammatical explanation :  a successful grammar is  a grammar that not only proposes coherent explanations of the raw data but one which succeeds in explaining satisfactorily the occurrences  left unexplained by  anterior grammars. Another  interesting case  in point is the behaviour of the verb RESEMBLE :

(20)a) Mary resembles her mother.

(b) Mary is resembling her mother more and more.

c)* Mary is resembling her mother. 

Both Martin Joos (1964) and  Ronald Langacker (1987) have tried to account for examples (20) : the first estimated that  BE+ING made the resemblance « more emphatic » ; as for Langacker , he takes up the same tautological explanation in his « Cognitive Grammar » (Vol.1, p.256) :

 9a) J.P.is resembling his father more and more every day.

 « RESEMBLE in (9a) is inchoative and describes an increase of similarity, in contrast to the constant degree of similarity imputed by (6a) : JP. resembles his father.  »

In my analysis, (20a) is a three - part utterance with a rhematic complementation (MOTHER is an open choice) whereas in 20 b) MOTHER is a closed choice (just as in I am leaving tomorrow, TOMORROW was a closed choice). The scope of the quantifier MORE AND MORE demands a binary relation where MOTHER is thematic.

            Now I intend to examine  the problem of the English DO,  which played an important part in the genesis of the Double Keyboard Theory and in the construction of my metaoperational grammar.

DO was always considered to be « a complication of the grammar of English » . This intruder, which baffled native speakers and foreign learners alike,  was exorcised by words which betrayed the embarrassment of descriptive grammarians (of course you could always cover up the problem with a historical veil !) : expletive or dummy.

In my « Esquisse d’une Théorie de DO » (Neuchatel 1974) I  claimed that DO, like –ING, was the trace of a processing operation, a metaoperator whose role was purely formal and whose presence  in the emphatic, negative and interrogative forms  revealed its raison d’être. DO was the trace on the surface of the predication operation , an operation which in most languages was a covert one. Hence the importance of that abstract witness of the  processing activity of the utterer , probably the only witness of its kind in the languages of the world (the German TUN is used metalinguistically …but only in dialect: “Ich tue deine Hilfe brauchen”). With DO we are  dealing with another  example of the natural metalanguage   which is at work in language. If  –ING  is a formal invariant signalling thematicity in the verbal group, DO is doing the same job at the level of the predication link. I am going to limit myself to an example of the so called emphatic utterances :

(21) He ‘did open the safe !

The emphatic stress bears on the tracer of the predication, that is DID . Contrary to traditional views , the emphatic element is not DID but the strong stress on DID ! In the negative form:

(22) He did not open the safe.

the negator quite logically bears on the link of the predication. Let us precise that the distance between the positive utterance : He opened the safe and its negative counterpart is far more important than  is currently believed. In the positive utterance the link of the predication is the verb OPEN whereas in the negative one it is DO (notice that they both carry the –ED mark of the preterite).

In my 1974 paper I reproached Chomsky with his ad hoc treatment of DO : in fact I do not think that the generative model was in a position to account for the emergence of DO (the only merit of Chomsky at the time was his global treatment of his DO-insertion rule !). Emonds’s DO-deletion rule brought nothing new except that it underlined the ad hoc procedures of generative grammar. The latest study of DO  is due to W.Hirtle . You will find it in the Jubilee hundredth issue of LINGUA (1997). The 38 - page paper, entitled « Do Auxiliary- A meaningful Support and Operator » is very disappointing : his treatment of DO as « a stretch of duration » does not hold water. I forgot to say that there is nothing new either in Chomsky’s Minimalist Program.

DO has to be studied contrastively  . The careful observation of the uses of BIEN in French or I in Polish in Russian  is illuminating. A contrastive analysis brings to light the fact that languages do not all present the same facility of access to the natural metalanguage : the me- talinguitic readability of languages  offers a wide field of investigation for the future. Meanwhile the English DO , with his extraordinary spectre of uses can be regarded as a sort of grand slam of linguistic analysis.

            With –ING and DO the grammar of English  offers two examples of thematicity : -ING concerns the predicate and DO signals the thematicity of the predicational  node . Anyway, in both cases we are dealing with  presupposed entities. From this point on the Double Keyboard falls into place : successively  it appears that the V1 to V2  /  V1 V-ING opposition  and   grammatical  pairs such as a / the, this / that, till / until, nearly /almost etc function according to the two-phase microsystem. Little by little an extraordinary symmetry emerges that replaces the « diabolical pairs » of English grammar  and the laundry-lists of descriptive grammar. The best I can do now is to comment on authentic utterances  - diagnostic examples -  which  have been chosen for their quality as  tokens of the cryptic system  that makes English understandable and learnable.


            Commentary of  some of the diagnostic examples  proposed in ‘CLEFS POUR BABEL » (EMA editions, 2000)  (pp.74-79)

            I have always been contrastive-minded in my linguistic research. In my thesis on BE+ING I systematically referred to French , Polish, Russian or German  when discussing  the problem of BE+ING.  In particular,the French distinction Passé Simple / Imparfait and the aspectual pair Perfective / Imperfective   could not be eluded  in my quest for an invariant for BE+ING. It so happened that my thesis opened up new vistas on the above mentioned questions. I  found that the Passé Simple/Imparfait opposition  was based on the same microsystem as the English opposition minus/plus BE+ING . In other words the Passé simple was the rhematic Phase 1 and the Imparfait the thematic Phase 2 : the two  past tenses of French grammar corresponded to the OPEN/CLOSED paradigm we have seen in the functioning of English grammar. Let us examine the following examples to make things clearer :

            a)Christophe Colomb découvrit l’Amérique alors qu’il cherchait la route des Indes.

 b) (  Inscription found in the Museum d’Histoire Naturelle in Paris) Météorite tombée               en 1492. Cette année-là Christophe Colomb découvrait l’Amérique.

 The  -IT / -AIT  opposition reminds us of the minus/plus BE+ING one : in both cases we have to deal with a ternary predication in the Passé Simple utterance (Christophe Colomb/ découvrir/l’Amérique) and with a binary relation in the Imparfait : Christophe Colomb  /  découvrir l’Amérique.

Here is another example which is self-explanatory :

c) Napoléon Bonaparte naquit à Ajaccio le 15 août 1769.

d) Six mois plus tôt il naissait Italien.

      Everybody knows that Slavonic languages manifest an original  partition of their verbs in their dictionaries : according to aspect, infinitive verbs can be perfective or imperfective. For example the verb WRITE receives two translations in a bilingual English-Polish dictionary :

            WRITE   :        pisać (Imperfective)

                                    napisać  (Perfective)

Let me add to make myself understood by  everybody that PISAć gives PISZę in the present (I write or I am writing) and PISAŁEM (fem.PISAŁAM) which corresponds to the English I WAS WRITING (the French Imparfait : j’écrivais). On the other hand NAPISZę is I will write and NAPISAŁEM translates the English preterite I WROTE or the French Passé Simple j’écrivis. Current grammars of Polish and Russian continue to present the Perfective /Imperfective opposition in spite of recent research in the field and  in spite of innumerable counter-examples (the same situation prevails in grammar-books of English as far as the BE+ING form is concerned).

What I have been claiming for years is that the Perfective/Imperfective opposition is in fact a manifestation of the two-phase microsystem. You can find a complete demonstration either in my « Clefs pour Babel » or in my 1995 paper. For want of space I shall propose only one Polish example to show  what is at stake :

1-     Wczoraj        spotkałem      Zosię na dworcu.(Yesterday I met Sophie at the station/ Hier je rencontrai S. à la gare).  

2-      Co niedzielę spotykałem Zosię na dworcu( .On Sundays I used to (would) meet Sophie at the station/ Tous les dimanches je rencontrais S. à la gare.)

The opposition SPOTKAŁEM / SPOTYKAŁEM  appears more neatly in French : je rencontrai vs. je rencontraiS. The analysis of example b) is parallel to that of French where TOUS LES DIMANCHES bears on the predication link –AIT, multiplying the relation by as many Sundays as you can count. The role of the English modal WOULD is also quite clear.

 Ample explanations on all the above points can also be found in my « Le Français Déchiffré, Clé du Langage et des Langues » (Armand Colin 1991) which opens with the microsystems VOICI/VOILA and A / DE .


The Double Keyboard Theory shows us that there is a covert symmetry between categories of grammar that have nothing in common at first sight. The common denominator of the this/that microsystem and the nearly/almost microsystem is to be looked for in the abstract two-phase system that regulates the emergence of BE+ING or DO. This cryptic basis functions on a very simple opposition : OPEN paradigm versus CLOSED paradigm. I have  only exposed the Double Keyboard of English  but I could have demonstrated that the same dichotomy explains the functioning of French  (see «  Clefs pour Babel » where full details are given) and of other languages (German, Portuguese, Polish). What all these analyses reveal is that the grammatical distinctions we find in our monumental grammars reduce to the ITERATIVE APPLICATION OF THAT BASIC DICHOTOMY (hence THE PRINCIPLE OF CYCLICITY). However surprising such a fact may appear it is a demonstrable fact. Here is a diagram showing the mechanism I am proposing :

                                                                    This  / that

            a / the                                                                                               to/-ING

 some /any                                             OPEN  /  CLOSED                                till / until 

              nearly / almost                                                                            O, -s, -ed/ do,does, did

                                                                  Shall /  will

 The first time I have used the label PRINCIPLE OF CYCLICITY was in my book about the acquisition of  the mother-tongue by children : « Caroline Grammairienne en Herbe ou Comment les Enfants inventent leur Langue Maternelle » (Presses de la Sorbonne Nouvelle 1995). In my opinion the key to L1 acquisition is  in the organization of language itself : languages are built in such a way that children cannot fail to discover their secret. The intelligence of the child meets the intelligence of the system . The Principle of Cyclicity ensures that the whole of language does not have to be learnt since any of the major grammatical problems can trigger all the others. In spite of my genuine admiration for Noam Chomsky whose trajectory I have been following since 1957, I do not agree with the  inneist theory and I consider Plato’s Problem  a fallacy.

            The Principle of Cyclicity  is also of paramount importance for the construction of pedagogical grammars . Learners of L2 will now find an organised presentation of the grammatical structure of the foreign language instead of  the unprincipled juxtaposition of grammatical items, based on a purely observational basis . The new approach will permit a programmed grammatical teaching  which was absolutely unthinkable in the descriptive model, with its artefacts and blind alleys. I shocked an assembly of teachers of English one day with a lecture entitled : « Is English learnable ? ». Well, English as a Foreign Language was not  only UNlearnable, but also UNteachable. On the level of linguistic research, things were not very much better : the descriptive approach  proved sterile in many doctoral dissertations . I have in mind a thesis on BE+ING published in the seventies whose conclusion claimed to be a confirmation of the” limited duration “theory and another very interesting study of the pair nearly and almost, with hundreds of excellent examples but which unfortunately went completely astray for lack of  linguistic insight .


 The main credo of metalinguistic grammar is expressed in the conviction that

            the linear utterance is not the message.

 Here is the linguistic credo of metaoperational theory  (« Clefs pour Babel » pp. 49-50 )

            The linear utterance is the final product of simultaneous/successive cipherings :

1-      The phonological ciphering :

The phonemes with which the speaker builds up the signifiers (« signifiants ») belong to  a closed system of oppositive units. In Saussure’s signs , signified and signifier are closely linked. Apart from the phonemic system stricto sensu there are rules of prosody which apply to the linear string (rhythm and intonation). Phonology seems to me to be the best introduction to linguistics : the systematic patterning of the phonemes as combinations of a handful of minimal distinctive features is a good stepping-stone to the systematicity of grammar.

2-      The lexical ciphering :

The utterer has inherited a lexical stock which is by no means a series of labels attributed to the objects of the extralinguistic world. Nothing will be said here about the possibility of a « mentalese » underlying and preceding the structuring of the utterances (see S.Pinker’s point of view in his well-known book : « The Language Instinct »).

3-      The syntactic ciphering :

The processing of the utterance consists in a series of deep abstract ordered operations.

On the grammatical level the utterer does not operate with ready-made chunks but can play with the potentialities of his grammar. In doing so he takes into account the situation of enunciation, the verbal context and the nature of the intended message.

            The morphology of the lexical elements (grammatical gender and number of nouns, verbal endings etc.) and the order of their appearance in the final linear chain, as well as the tracers of the structuring operations (free or bound morphemes such as a/the, this/that, to/-ING, do, un(e)/le, la, à/de, bien ...) are the surface tokens of the ciphering operations executed by the utterer/architect. The result of this complex processing (Humboldt’s ergon) represents an algebra which mirrors in a revealing way the encoding of the utterance, and not the extralinguistic world, hence the impossibility of a direct assignment of  the meaning. Therefore one can understand easily the totally artefactual character of an analysis which tries to assign meaning to the successive elements of the linear string. However this is exactly what has always been done and what is still being done in descriptive grammar. This atomistic approach has produced catastrophic results both in the conception of grammar treatises and … in the brains of the learners ! How can one account for the use of a BE+ING construction by focussing on the verb alone ? What image can one give of the deep function of the French Imparfait if one does not see what sort of predication is being signalled by the –AIT metaoperator ?

Languages offer varying degrees of readability (lisibilité) in their surface utterances. For example we have seen that the English DO reveals a major metalinguistic operation of predication. In German, the order of elements in subordinate clauses, far from being a national freak, conveys precious information on the organisation of the sentence as a whole. In French the à/de microsystem can almost be deciphered at surface level. What must be kept in mind is the fact that a particular grammatical point can be a cryptotype (cf. B.L. Whorf) in one language and almost a transparent marker in another. This explains why it is impossible to describe a language in isolation : autarcic grammars  belong to an outdated practice. Contrastivity is thus an absolute necessity in linguistic analysis.                                                                                       In my first grammar of English published in 1982 : « Grammaire Linguistique de l’Anglais » I tried to apply the principles enumerated above. It was the first grammar built on an original linguistic theory and it differed totally from the grammar-books in circulation at the time.First there was a preface where I presented my approach to grammar and where I defined a certain number of prerequisites . So far grammar-book authors did not bother to explain their conception of grammar to their readers . Second, each chapter began with an introduction explaining  not only the contents of its pages but the precise aim of that chapter in the construction of the totality of the grammar. For exemple, the introduction of the first chapter (entitled : « La Relation Prédicative : Sujet- TO – Prédicat et les Opérateurs O, TO , -ING ») warned the reader that he would be dealing with a grammar where the sequencing of the chapters was neither due to chance nor to a whim of the author :

            « Le premier chapitre d’une grammaire ne peut en aucun cas devoir sa qualité de premier au seul fait qu’il serait placé avant le deuxième. Il doit être premier au sens fort, c’est à dire qu’il doit conditionner la suite, constituer la base sur laquelle on va construire les autres chapitres ». (p.9)

            « The first chapter of a grammar should in no case owe its first place to the fact that it is situated before the second. It must be first in a strong sense, that is to say that it must condition what comes next, constitute the basis upon which the other chapters will be built ».

What does all this mean ? It means that to deserve the label GRAMMAR the presentation of the functioning of a language must be a coherent whole , not a hodge-podge of unrelated problems. Metaoperational grammar was founded on the DoubleKeyboard theory and so was in a position to present an organised, programmed grammar. For the first time in the history of grammar the choice of the starting-point of the grammatical enterprise was crucial . Why did I decide to begin my book with the TO/-ING opposition in the first place ? My aim was to kill two birds with one stone : first give my reader an idea of the rhematic/thematic microsystem  in the functioning of V1V2 constructions (V1 TO V2 vs V1V2-ING) –and this was a completely new approach to English grammar- and then  pave the way to chapter two where the  concept of presupposition  would be essential to undertand the grammar of the BE+ING constructions.The Double Keyboard Theory  was going to accompany the reader  throughout the whole grammar : the modals, DO, a/the, this/that, who, which/that etc.

I could have adopted a different strategy and start with the metaoperator DO : this was also an excellent opportunity of acquainting my readers with the notion of tracer of an operation and simultaneously prepare them for more metaoperators of the natural metalanguage of English. Another opening had tempted me for some time : why not place the chapter on the Modals first ? This would have enabled me to show the mirage of linearity and the geometry of the grammatical elements tradition had  misnamed “defectives » ! Be it as it may, when you are introducing novelty into a field of knowledge, you cannot  succeed if you do not help your readers to forge the tools of analysis  that will be necessary to  analyse the raw data of language. I understood that simple truth still better when composing my second grammar-book, ten years after the first : « Les Clés de la Grammaire Anglaise » (The Keys to English Grammar). This time, I placed a list of twelve Preliminaries in a separate introductory chapter that the student was expected to read BEFORE tackling the grammar itself.

            Grammar is not a bed of roses. My generation has witnessed the birth and the decline of many attempts at capturing  language . The short-lived  epistemological  reign of so many successive approaches has provoked  prudence and suspicion in the ranks of researchers and teachers. This state of things explains the success of  eclecticism : let’s pick and choose the best achievements of the different theories which are to be found on the market. Alas ! An eclectic approach to the grammar means the death of grammar  since eclecticism is the negation of coherence and systematicity, the negation of what makes language language.

            To conclude this rapid explicitation of metaoperational grammar I would like to insist on the importance of L1 in pedagogical grammars of English. I have already remarked the impossibility of autarcic grammars : no language is able to explain its working  by his own means.  Pedagogical grammars of English should be based on the grammar of the L1 for the simple reason that the grammar of L1 is the best port-hole (hublot) through which the learners can apprehend language.We all know how important language-awareness is in the learning of languages. And we also know how difficult it is to convince  the speakers of a language that being able to speak a language  does not make you an authority  to speak ABOUT it. Here are the opinions on that painful subject of two famous linguists :

Benjamin Lee WHORF : « The truth that those who easily and fluently use the intricate systems of language are utterly blind and deaf to the very existence of those systems, until the latter have been, not without difficulty, pointed out ».(« Language , Thought and  Reality », MIT 1956,p.256.

Gustave GUILLAUME : “La langue est pleine d’opérations mystérieuses(…) L’étude de la langue nous met en présence de choses auxquelles un esprit même cultivé peut n’avoir pas songé, et qui s’écartent complètement des voies suivies par la pensée courante ».(« Leçons de Linguistique, 1945-46 A , N° 7, 1987).

( A language if full of mysterious operations(…). The study of a language confronts us with things that even a learned mind may not have dreamt of and which are completely foreign to current thinking).


            I have said nothing on  a very important subject : the quest of Universal Grammar . Noam Chomsky and his followers have lately devoted  quite a lot of space on this  theme, especially in the latest developments of generative theory. U G plays an important role in the problem of the acquisition of L1 by children. However interesting all those developments may be, I for myself have not been convinced by the arguments  and the linguistic examples put forward to defend the inneist stand.

            I claim that my Principle of Cyclicity with its underlying Double Keyboard Theory is a  step towards a different conception of Universal Grammar. My translingual invariants  , which have emerged  in my research on  the comparison of languages,  tend to show  that similar operations of utterance processing can give quite different surface traces : the English DO, the French BIEN , the Polish and Russian I or the Turkish –DIR  do have something in common, however diverse their particular spectres may be. The same is true for the ENGLISH BE+ING, the French Imparfait or the so-called Imperfective Aspect in the Slavonic languages.

My last  word will be about the teaching of English as a Foreign Language. The current presentations of English grammar make English unlearnable . Everybody agrees that the teaching of English, at least in France, is a failure, to say the least (some even speak of “disaster”) . However, any novelty in the grammatical approach of the language of Shakespeare is considered as an outrage (by publishers, teachers and students !).So the old drudgery continues in spite of the dozens of theses and hundreds of papers   offering  new  viewpoints to remedy  the present situation. Only the end of the resistance of the old paradigm  will signify the end of the blindman’ s buff. Meanwhile linguists will continue to be spurred on by “the wanting to see the wheels go round”.


Henri Adamczewski

Professeur Emérite à l’Institut du Monde Anglophone 

Université  de la  SORBONNE  NOUVELLE


 Bibliography : 


-         "BE+ING dans la Grammaire de l'Anglais Contemporain", 720 pages, thèse d'état 1976, Champion Paris 1978.

-         "Grammaire Linguistique de l'Anglais", Armand Colin 1982.

-         " Les Clés de la Grammaire Anglaise", Armand Colin 1991.

-         "Le Français Déchiffré, Clé du Langage et des Langues", Armand Colin 1991.

-         "Caroline Grammairienne ou Comment les Enfants inventent leur Langue Maternelle", Presses de la Sorbonne Nouvelle, 1995.

-         "Clefs pour Babel ou la Passion des Langues" , Ema, 60340 St Leu d'Esserent, 1999-2000.

-         "La Problématique de l'Aspect en Français et en Polonais. Une Nouvelle Approche: la Théorie des Phases" in Zofia Cygal-Krupa Ed. "Les Contacts Linguistiques franco-polonais", Presses Universitaires de Lille, 1995.

 CHOMSKY, Noam :

"Knowledge of Language, its Nature, Origin and Use" , Praeger, New York, 1986.

 JOOS, Martin :

Review of A.Ota's "Tense and Aspect in Present-Day American-English ". LANGUAGE VOL.40 N°3, 1964. 

HIRTLE, William :

"DO Auxiliary. A Meaningful Support and Operator". LINGUA N° 100, 1998.

 LANGACKER, Ronald :

"Foundations of Cognitive Grammar", Vol.1,Stanford University Press 1987. 

PINKER, Steven :

"The Language Instinct", Penguin 1994.

 WILMET, Marc :

"Grammaire Critique du Français", Duculot 1998.