¹ Romania, xlvi, 46–83. In a later article on another subject, Foulet recapitulates, and but slightly modifies, his theory of 1920 (Romania, lxii [1936], 43–51). He also suggests (pp. 51–55) that his same hypothesis should be accepted for English.

1a In OF, as M. Foulet points out, ce regularly preceded the verb (at least in declarative sentences), whether its function was that of subject (c'est vrai) or of complement (ce fait il). Hence, the ambiguity of c'est il.

² The first appearances of the new construction are attested by Foulet (in 1920) as follows:

(p. 58) c'est moi 1371–72 (Lelivre du Chevalier de la Tour Landry)

(p. 66) c'est lui 1455–60 (Petit Jean de Saintrê)

(p. 62) c'est vous 1456–61 (Cent nouvelles nouvelles)

(p. 69) ce sont eux 1461–68 (Le jouvencel)

(p. 69) c'est toi

(p. 69) c'est nous Not before 16th century; found by Foulet only in Meigret.

(p. 69) c'est eux

In 1936 he amends this chronology somewhat by adducing earlier examples (in both cases, more than a century earlier) for the two forms c'est vous and c'est lui. His attention had been called by Armstrong to an example of c'est vous in a 14th-century MS of Y pomedon (probably first half of 14th c.); and on the basis of this, he decides to accept as a genuine case of c'est lui an example from the Miracles de Nostre Dame (c. 1350) which he had already cited in his earlier article, but had refused to accept (see below, note 39).

³ As for the twofold construction represented by the third person plural, M. Foulet gives several pages (pp. 71–74) to a discussion of the career (from the 16th century on) of the “logical” c'est eux, and the compromise form, illogical but “correct”, ce sont eux.

⁴ That is to say, of the first part of the article, in which M. Foulet gives his explanation of the development from ce suis je to c'est moi. In the last part, after concluding the demonstration of his theory, M. Foulet offers examples of each of the new forms that he has been able to attest, following, in a general way, their gradual acceptance (particularly by grammarians); he takes up the much-debated problems of c'est eux vs. ce sont eux, and of c'est moi qui suis … vs. c'est moi qui est … —including also, along the way, a discussion of the grammaticalization of the interrogative pattern qui est-ce qui …?. Many of his remarks in the final part are, if somewhat cursory, interesting and informative, their validity in no way dependent on the tenability of his main thesis.

⁵ It is true that c'est eux is the popular construction. Nevertheless, no one would maintain that cultured persons who prefer ce sont eux have no feeling for their language. The truth is that, in such a predication, the verb is subject to influences coming from two different directions, and the triumph of either subject or predicative can be justified. The very fact of constructing such a bold equation as “A is B” must constitute a (potential) challenge to the rules of normal syntax; and it is not surprising if, occasionally, the fundamental law of the agreement of subject and verb is broken. Indeed it is possible for the form of the predicate noun to influence not only the verb but the subject itself, which is assimilated (in gender) to the predicative. In Classical (and pre-Classical) Latin there was a strong tendency to transform a neuter subject (a conceptual neuter) in accordance with the gender of the predicative: haec est stultitia instead of id est stultitia (cf. Stolz-Schmalz, 210). In later Latin the type haec est stultitia became the exception.

⁶ Cf. also Germ, das sind meine Freunde as well as es sind zwei Ritter gekommen (here French distinguishes itself from German: il vint deux chevaliers).

⁷ We shall see later that ∗c'est ils was not, after all, the form logically to be expected.

⁸ M. Foulet does not, however, explain the absence of ∗c'est ils as (exactly) due to a desire to anticipate the predicative; he seems to think the reason lay in the subject form of ils. Even though this would have been felt as predicative (in the hypothetical ∗c'est ils), still its form would have been reminiscent of the subject pronoun, so that the sequence … est ils would have been impossible. This objection could not, of course, apply to c'est eux, where the oblique form is used for the predicate. Nevertheless, M. Foulet maintains that c'est eux, too, was a problem; he believes that it must have met with very strong opposition at the beginning, since it was so late making its appearance in literature (16th c.) But why should there have been the slightest difficulty (and M. Foulet, who has overlooked an example of c'est eux in the Cent nouvelles nouvelles, may be exaggerating somewhat the case against this form) in the way of accepting c'est eux, where the form eux is clearly not that of subject—unless there were, indeed, a desire to anticipate the predicate? Not once (even when he cites sa maladie sont des vapeurs) does M. Foulet mention, as a general principle operating in language, this tendency of anticipation.

⁹ Abbreviations used in referring to the OF texts cited in this article:

Adams. Das Adamsspiel. ed. Grass (Halle, 1891).

Anc.Th. Ancien théâtre français, ed. Viollet-le-Duc (Paris, 1854), Vols. i-iii.

A+N Aucassin et Nicolette, ed. Roques (Paris, 1925).

Ber. Bérinus, ed. Bossuat (Paris, 1932), Vol. ii.

Berte Li roumans de Berte aus grans piés by Adenés li Rois, ed. Scheler (Brussels, 1874).

CNN Les cent nouvelles nouvelles, ed. Champion (Paris, 1938).

CR La chanson de Roland, ed. Bédier (Paris 1937).

CV La chastelaine de Vergi, ed. Raynaud (Paris 1921).

Cont. La continuation de Perceval by Gerbert de Montreuil, ed. Williams (Paris, 1922).

Eli. Eliduc in Les lais de Marie de France, ed. Warnke (Halle, 1925).

En. Eneas, ed. Salverdade Grave (Paris, 1925), Vol. i, 1929; Vol. ii.

Esc. L'escoufle, ed. Michelant and Meyer (Paris, 1894).

Fab. Recueil … des Fabliaux des XIII^e et XIV^e siècles, ed. Montaiglon and Raynaud, Vols, i, ii, (Paris, 1876–77).

F+B Li romanz de Floire et Blancheflor, ed. F. Krüger (Berlin, 1938).

Fol.Tr. La folie Tristan (MSS. Oxford and Berne), ed. Bédier (Paris, 1907).

G+A Le garçon et l'aveugle, ed. Roques (Paris, 1912).

Gers. Ad Deum Vadit by Jean Gerson, ed. Carnahan (Univ. of Illinois, 1917).

Greb. Le mystère de la passion d'Arnold Greban, ed. Paris and Renaud (Paris, 1878).

Guig. Guigemar, in Les lais de Marie de France.

G.d'A Guillaume d'Angleterre by Chrétien de Troyes, ed. Foerster (Halle, 1899).

HdeB Huon de Bordeaux, ed. Guessard (Paris, 1860).

JdeF Le jeu de la feuillée par Adam le Bossu, ed. Langlois (Paris, 1911).

Lanc. Lancelot by Chrétien de Troyes, ed. Foerster (Halle, 1889).

Lanv. Lanval, in Les lais de Marie de France.

Mel. Mélusine by Jean d'Arras, ed. Stouff (Paris, 1932).

Mir. Miracles de Nostre Dame, ed. Paris and Robert (Paris, 1876–83), Vols. i-vii.

Nic. L'évangile de Nicodlme, ed. Paris and Bos (Paris, 1885), 3 versions: A, B, C.

Nouv. Nouvelles françoises en prose du XIII^e siècle (Paris, 1856); du XIV^e siècle (Paris, 1858), ed. Moland and D'Héricault.

Path. Maître Pierre Pathelin, ed. Dimier (Paris, 1931).

RdeC Raoul de Cambrai, ed. Meyer and Longnon (Paris, 1882).

Ren. Le roman de Renart, ed. Martin, Vol. i (Paris, 1882).

R+M Le jeu de Robin et Marion by Adam le Bossu, ed. Langlois (Paris, 1924).

Ruteb. Rutebeuf, Oeuvres, Vol. ii, ed. Jubinal (Paris, 1874).

SCSC Mystère de St. Crespin et St. Crespinien, ed. Dessalles and Chabaille (Paris, 1886).

St. A. Saint Alexis, ed. Storey (Paris, 1924).

St.L. Saint Léger, ed. Linskill (Paris, 1937).

Wist. Wistasse le Moine, ed. Foerster and Trost (Halle, 1891).

Yv. Yvain by Chrétien de Troyes, ed. Foerster (Halle, 1891).

¹⁰ In ce n'est il pas? the (negative) declarative form is used with interrogative force, perhaps because n'est il pas ce? was felt as slightly awkward (ce being so far from its normal position at the head of the clause).

¹¹ I have been unable to find an example of est il ce? in the 12th or 13th centuries, the period during which the old pattern still retained its original interpretation; the interrogative form of our pattern was very rare in the third person. However, this is one case in which it is quite legitimate to cite a later example as evidence for the early period: if ce was still felt as predicative (contrary to the modern feeling) as late as the 15th century, we may be sure that this was its function in the beginning.

¹² It may become apparent later why M. Foulet failed to use either of these arguments in his interpretation of c'est il.

¹³ Compare, for example, “il lui baisa la main” vs. “il baisa la main de la jeune fille.”

tu (late 15th century?)—which, in turn, by way of c'est vous, c'est nous, ∗c'est je, was to lead to c'est moi (16th century?). But, as we have seen, c'est moi is found in 1371—that is, the new construction, full-fledged, supposedly representing a half-dozen stages after the “new” c'est Jean, appears before the “lutte acharnée” against the “old” c'est Jean had begun !

And the confusion is only deepened by M. Foulet's additional remarks in 1936; having finally admitted c'est lui for 1350, and bound by the terms of his theory to postulate c'est moi before c'est lui, he even goes so far (by a process of reasoning which I cannot follow) as to suggest the possibility of a c'est moi “dès le XIII^e siècle” (p. 50) ! This would of course put back even earlier the reinterpretation of c'est il, and still further back that of c'est Jean (12th century?). Apparently Foulet has completely forgotten the “lutte acharnée” at the end of the 14th century, to which he originally attributed the reinterpretation of c'est Jean. Everything now is dissolved in the haze of rhetorical questions: “A quel moment l'ordre des mots, tel que nous le connaissons aujourd'hui, s'est il imposé à la langue, et a-t-il cessé d‘être un ordre fréquent, puis un ordre favori, pour devenir l'ordre unique? En d'autre termes, à quel moment la ruine de la déclinaison a-t-elle modifié nécessairement l‘économie de la phrase française? Toutes questions délicates … ”(p.50).

¹⁶ Unless one should consider as belonging to the type c'est de Jean que …, such an example as “Est ce a certes ou a gas/Que li rois n'i entrera pas?” (Ren., xi, 2501–02), where the predicate prepositional phrase is used adverbially, or adjectively (cf. also JdeF, 132).

¹⁶ It is difficult to say just how old c'est Jean (with its neuter pronoun; is. We will not, so far as I know, find ∗hoc est Johannes in Latin (at least, not in the function here in question); only hic est, ille est Johannes. However, though c'est Jean is not a direct continuation of hic est Johannes, still, both represent the same general type of predication: identification of an unknown person (and the relationship of subject and predicate must be the same in both). But as concerns the periphrastic prepositional type c'est de Jean que …, there is nothing remotely comparable to this in Latin.

¹⁷ The only case in which c'est Jean would be close to c'est de Jean que …, is when the former is followed by a limiting relative clause depending on a preposition: c'est Jean dont je parle (vs. c'est de Jean que je parle). Indeed, in late OF a confusion between these two types was to arise: cf. “c'est a vous a qui je vendy siz aulnes de drap”, Path., 1265–66). But in the early period the type (with relative pronoun) c'est Jean dont …, and even c'est Jean qui … was exceedingly rare (just about as rare as the type c'est de Jean que … itself). See below, n. 55.

¹⁸ And now we see why M. Foulet did not, at the beginning of his article, use the test of interrogative (etc.) word-order in his explanation of the original interpretation of c'est il. If he had begun by offering est il ce?, ce n'est il pas?, as the acid test for the function of ce, he would have been expected to use the same test on c'est Jean, where the evidence would have been contrary to his theory of ce-predicative. And it is much more difficult to explain away evidence in one case if it has just been accepted as definitive in a preceding case. Accordingly, M. Foulet preferred to establish the interpretation of c'est Jean without any evidence, as needing no proof. To slip in a conclusion as a self-evident fact (“thus, ce is the predicative in c'est il—and the same, of course, will be true of c'est Jean”) is often the easiest way to have it accepted, implicitly and uncritically, by the reader. Then, after c'est Jean has been accepted as containing ce-predicative—only then is the factor of word-order mentioned at all (and surely not as an acid-test): now M. Foulet (within the established framework of ce-predicative) can afford to bring in an “apparent” inconsistency in word-order. And since word-order had not been mentioned before, this admission appears as a magnanimous concession; the inconsistency itself is accepted as not too important (for is it not “self-evident” that ce is predicative?), though puzzling. And the ingenious solution of this puzzle which M. Foulet offers, is gratefully accepted by the reader, who does not see that the “puzzle” is nothing less than absolute proof against M. Foulet's initial assumption.

¹⁸ In the formulas with cil and cist we have, of course, a direct continuation of the Latin type with personal demonstrative, whereas c'est Jean shows the substitution of the neuter (∗hoc est Johannes). How and when did the de-personalized type *hoc est Johannes arise? So far as I know, no study has been made of this extension of the neuter pronoun in Romance (the same phenomenon, of course, exists in many other languages). One may, in a post hoc fashion, explain this use of the neuter in identification by the abstract nature of the unknown: an entity standing in need of identification may be more easily denied concrete existence. Generally speaking, the type c'est Jean serves one of two functions: (A) to identify a physical presence and (B) to define more specifically a name given by the context:

(A) Atant es vos Brun a la hese. / “Renart” fail it, “parlez a moi!

Ge sui Brun messagier le roi …“. / Renart set bien que c'est li ors,

Reconneü l'avoit au cors. Ren., i, 490–495

(B) “Enseigne moi le plus riche homme / De ceste ville …”

“Dist li vilains: / ”C'est nostre Prestres.“ Fab., ii, 49

In the first case, where ce refers to a presence, we may imagine that the use of the neuter arose in situations where the one to be identified is first perceived as a dim shape, or, perhaps, makes his presence known only by a noise: cf. “Dedenz commence a regarder/Et son ombre a aboeter. / Cuida que ce fust Hermeline / Sa femme” (Ren. iii, 159); here the unknown person exists first of all as a stimulus to the senses, a phenomenon—an abstraction. In the second case we have an abstraction of another sort: when ce refers to a name in the context, it is possible to think of it as referring not to a person but to a “label.” It may be that this type is ultimately connected with the Latin formulas id est …, hoc est … (orginally used only parenthetically); cf. “Quant anfans fud … / al rei lo duistrent soi parent, / qui donc regnevet a ciel di; / cio fud Luthiers, fils Naldequi” (St.L., 14–16). In OF we find the two types, cil (cist)+pred. noun and c'est+pred.noun alternating in much the same situations; it may be said, however, that the personal demonstratives served rather the function of “presentation”, the neuter that of “identification”, though the latter was not limited to this function; indeed, ce could even be used occasionally (as it frequently is in Modern French) to characterize a person already known: “de ma bielle fille vous sai je bien à dire boines nouvielles; car je l'ai veue maintenant, et si saciés ke c'est la plus bielle dame ki soit el monde” (Nouv., xiii, 143).

²⁰ According to my examples, the type … est Jean is only to be found in dependent clauses after verbs of thinking, seeming, etc.

²¹ The oblique form celui was used not only as direct object but also as predicative: not exclusively, in this last function, but more frequently than was the nominative (alone possible in Latin). As for cist, cestui, this pronoun was rarely used as predicative.

²² The subject-less construction ce suis, ce est rarely appeared without the intensifying particle mon. For the etymology of this particle, as well as for its use in OF, see the article of Professor Spitzer, “La particule mon”, PMLA, lxi (1946), 607–620.

²³ That in such questions as “who is that?”, who must be the predicative and that the subject, seems to be generally recognized: cf. Jespersen (The Philosophy of Grammar, p. 153), who contrasts this type with “who is ill”?, where who must be the subject. The difference between the two is shown by the word-order in indirect questions (the same in both French and English): “he asked who that was” but “he asked who was ill.” Again, in questions of identification, the verb regularly agrees with the second pronoun: “who are you?”, “who am I?”, “who is he [that]?”. Even Foulet agrees (p. 52) that in “qui est-ce?”, ce is the subject; oddly enough, he uses this fact as evidence for the subject function of ce in c'est de Jean que …, but not in c'est Jean.

²⁴ The interpretation “John is that” would, of course, be in accord with logical sequence, if the predication were used to answer some such questions as “Is John your husband?” (here, in both cases. John would be the subject). But I have never found such a sequence as “Est Jean votre mari?”—“Ce est Jean”, for the obvious reason that, if the name were used in the question, it would hardly be repeated in the answer; we would find, instead, c'est mon, si est, etc.

²⁵ It might be asked: if logic requires that ce be the subject in c'est Jean, what of the pronominal type ce suis je, c'est il, where ce must be the predicative? There are, theoretically, two explanations possible for the difference in function of ce in the noun type and the pronoun type respectively: either the latter is to be found in a different context, or else, if the context is the same, there must exist some special reason (not present with the noun type) for disregarding logical sequence. We shall have occasion, to see, later, that both explanations are valid for the pronominal type.

²⁶ It should be stated, in the interests of literal accuracy, that M. Foulet does not make the specific assertion that the reinterpretation of c'est il was influenced by that of c'est Jean.

He simply described, first (and in great detail), the shift that took place in the latter, because of the three reasons noted—and then continues by saying: “Il est clair que la locution c'est il ne peut résister au mouvement puissant qui emporte la langue. Ici aussi [!] ce sera désormais le sujet, il l'attribut” (p. 54). But it is impossible not to assume that he believed in this influence. Otherwise there would have been no point in beginning with an elaborate description of the checquered career of the noun type, in a paper dealing with the problem, “Comment on est passé de ce suis je à c'est moi.”

²⁷ Compare, however, note 14.

²⁸ A form never found, so far as I know, when identification was involved.

²⁹ As, indeed, it occasionally was in OF; cf. above, p. 1058. As a matter of fact, it is still possible to find in modern French je le suis (il l'est, vous l'êtes,) used in identification instead of c'est moi—for example, as a categorical answer to the question “are you … ?” (“Etes-vous la femme du prisonnier?”—“Je le suis”.) M. Foulet speaks throughout as if c'est moi were the only modern continuation of ce suis je, completely disregarding je le suis. This is because he never analyzes the meaning of the two constructions and the different contexts in which they may be used.

³⁰ That is, the fluctuating function of ce, and the loss of nominative -s. The first of these, as we have seen, did not exist. And as for the second, it may be stated here, somewhat belatedly, that this could hardly have played the rôle which Foulet assigns it. In the first place the -s of the original c'est Jehans could be no sure indication of subject function, since the nominative form served for predicative as well as for subject: “Margariz est mult vaillant chevalers” (CR, 1311). On the other hand, it is true (though, of this, M. Foulet, says nothing) that we find (already in the Roland) the nominative Jehans less frequently in the type c'est J. than in Jehans est. … Accordingly, if Foulet does want to see in the form with -s an indication of subject function, he must then explain the frequency of the form without -s, c'est Jehan, in the early period. If this means anything, it means that, in this pattern, the second substantive was felt as predicative, and that the loss of the -s was a result of this interpretation (in line with the preference for c'est celui over c'est cil), not the reverse.

³¹ I must confess that I do not quite understand this reasoning: does Foulet believe, perhaps, that Gerson, while feeling ce to be predicative (according to the traditional interpretation), would construct it as subject, simply in order to avoid awkwardness? But the language had already found a way to avoid the awkward n'est il pas ce?, and without upsetting sentence structure; as we have already seen (note 10) one simply used the declarative form with interrogative force: ce n'est il pas? And Gerson might have done the same. As a matter of fact, this is very probably what he did. For Foulet's citation (borrowed from Petit de Julleville) is taken from an unedited MS rejected by the editors of the (1917) critical edition of Gerson's sermon Ad Deum Vadit (cf. Bibliography). In this edition we find precisely: “Approuchez cy, ce n'est il pas?”

³² M. Foulet might have offered a much more reasonable explanation for the rarity of the form he wanted to find (assuming, of course, that such a form had been created), and this is, simply, the infrequency (already mentioned) of the interrogative form in the third person. In 46 texts I have found only 2 examples of est il ce?, as against 84 of c'est il. (In the second person, on the other hand, the interrogative es tu ce?, estes vous ce? was extremely frequent.)

³³ It should be stated that my “thirdly” is M. Foulet's “firstly”: in assembling his evidence for misinterpretation within the pronominal paradigm, he begins not with the third person where this is supposed to have originated, but with the second person (c'est[t]tu), where we are meant to see a further stage in the process of misunderstanding (which he has not been able to attest at its source).

³⁴ How, then, is es ce tu? to be explained? For if we refuse to admit both misinterpretation and misspelling in the same example (which is also Foulet's sole example for the fourteenth century), and insist on interpreting es as it stands (i.e., as agreeing with tu), then we are faced with abnormal word-order: es ce tu instead of es tu ce. But it is not impossible to find other cases of ce-complement out of position. We must remember that this weak demonstrative was half-way between personal pronoun and demonstrative; accordingly, it is not surprising that we may find ce occupying now the position of le, now that of cel, celui. This was particularly frequent in dependent clauses: for example, in the stock expression, “when so and so heard this” we may find the three arrangements: quant ce oï Jean (normal word-Order); quant Jean ce oï (ce in position of le); quant Jean oï ce (ce in position of cel, celui). In the two examples below we have obviously ce in the position of le:

Se vos çou faissiés (A+N, xl)

Quant vos ce dites (RdeC, 5281)

That ce was particularly susceptible to shifting word-order in dependent clauses is easily enough explained. In declarative (independent) sentences, the normal position of ce was at the head of the sentence—a position no longer possible when the clause is introduced by a conjunction; once uprooted, then, from its fixed place, ce could float hither and yon in the sentence.

Now, in interrogative sentences, ce-complement normally took end-position: fait il ce? est il ce? es tu ce?, etc. But the interrogative type was comparatively infrequent, and it could easily be understood if, in such cases, this end-position was felt to be slightly awkward; we have already seen that the (still more cumbersome) arrangement n'est il pas ce? could be replaced by the declarative formula (with interrogative force) ce n'est il pas?. A second solution would be for ce to assume the position of le: that is, on the model of fait le il? (which shows the normal interrogative word-order for le) we might find fait ce il? I have never found the exact type fait ce il?; but I have come across the rather unusual interrogative arrangement (with compound verb and noun subject): A ce fait Jhesus? (Greb., 12519), which surely shows a desire to avoid end-position for the weak ce. And so I suggest that in es ce tu? we have simply ce in the position of le: es ce tu? = es le tu? (like fais le tu?, vois le tu?, etc.). Furthermore, I would apply this same interpretation to the example of Gerson, above, which I temporarily dismissed as “hybrid”: n'est ce il pas? If we accept this example as a second possible reading (in spite of the somewhat suspect MS in which it is found), then we would have both solutions ([1] declarative word-order; [2] ce in position of le) applied to the same formula: n'est il pas ce?>1. ce n'est il pas? (reading in 1917 ed.)

>2. n'est ce il pas? (reading acc. to Julleville)

As evidence that this second solution came to be more and more adopted, at least in our formulas of identification, I need only point to some examples from Meigret, which Foulet himself quotes but fails to understand: we find with him both es ce tu? and [ne]seras ce tu pas. … Since this 16th century phoneticist spells es ce tu as esse tu, this one example, alone, might be questionable (i.e. = est ce lu, as Foulet wants to believe). But in the interrogative-negative there is double evidence against such an interpretation: (l)the form seras (is this, too, a misspelling?); (2) the word-order … tu pas; both are proof that tu must be the subject. Yet Foulet completely disregards, not only the spelling seras (as he did in his example es ce tu), but the final position of pas (as he did in the example of Gerson: n'est ce il pas); he treats Meigret's example exactly as if it read “ne sera ce pas tu”! I cannot stress too much the fact that in all three cases of the abnormal word-order of ce, ambiguity has been avoided: in es ce lu? by the agreement of the verb (which Foulet would destroy); in n'est ce il pas?, by the position of pas (which Foulet overlooked); and in ne seras ce lu pas? by both together! Language is, mainly, not ambiguous. And when the speakers seek to correct its occasional awkwardness, they are often able to do this in a more intelligent way than by “misunderstanding.”

³⁵ M. Foulet does not always recognize this. On p. 54, speaking of the “necessity” of the reinterpretation of c'est il, he remarks, “Il par sa forme [!] se prêtait à jouer l'un ou l'autre des deux rôles [subject or predicative]”, though, later, he will contradict this in his remarks on the non-existent *c'esl ils (see above). At any rate, it is clear why M. Foulet did not wish to use the subject form of il as evidence, in the beginning, for the original interpretation of c'est il.

³⁶ Cf. my article “Le type timbre-poste” (Word, ii [1946], 216–228), in which I oppose the opinion of Darmesteter that the modern type of elliptical compound (timbre-poste, papier-ministre) was due simply to a misunderstanding of appositional compounds: i.e., roman feuilleton “roman qui est un feuilleton” would come to mean “roman qui se trouve dans un feuilleton” (= *roman de feuilleton). In this case, indeed, the advocate of “misunderstanding” does not even consider it necessary to imagine some particular factor which might have led a 19th-century Frenchman suddenly to misunderstand a basic construction of his language.

³⁷ According to Foulet, Maupas, in 1625, still admits ce suis je along with c'est moi. It i^S very likely that ce suis je was the last representative of the old construction to disappear. In Provençal where, just as in French, the new construction is in general use, the old type still lingers in the first person singular: siéu iéu = [ce] suis je (cf. Ronjat, Grammaire Istorique des parlers provençaux modernes [Montpellier, 1937], p. 529).

³⁸ Compare his words, (p. 54), when he speaks of the irresistible movement which was to transform the interpretation of c'est il: “Il va de soi qu'à aucun moment, pas plus que dans le cas précédent [c'est Jean], on ne s'est rendu compte de ce renversement de point de vue. Il par sa forme se prêtait à jouer l'un ou l'autre des deux rôles, comme on voulait. A la faveur de cette indétermination, on a passé d'une conception à l'autre, au moment favorable, sans même s'en douter.” Andin a later reference (p. 68) to this same phenomenon, he insists again: “Il n'est même pas probable, comme nous l'avons dit, qu'on ait jamais eu conscience de ce renversement des rôles.”

³⁹ This example (from the Miracles de N.D.) reads as follows:

Or me dites, damoiselle Anne,

Cel home la, se Dieu vous sault,

Ressemble il bien a Musehault

Le messagier?

—Mais dites c'est sanz mençongier

Li proprement, (xxxvii, v. 867–871, t. vii)

And it is rejected by Foulet since, according to him, the author was “forced” to use the new construction (“c'est presque un cas de force majeure”: p. 65) because of the unusual, emphatic word-order involving enjambement. As if the author himself had not chosen this unusual word-order.

One is tempted to wonder if the real reason for Foulet's rejection of this example was not simply the fact that the date of its appearance is in contradiction with the order he postulates. As a matter of fact, he admits as much in 1936; for it is, apparently, the presence of an earlier c'est vous, discovered by Armstrong, which has forced him to “accept” the example of c'est lui.

⁴⁰ Foulet does mention an example (Rom., lxii, 48) of c'est moi meïsmes in the 13th century (“cité par M. Humphreys”), but this he rightly rejects as having nothing to do with the formulas of identification in question; the passage is taken from Méraugis (4877–90), and belongs to a monologue in which the hero is describing his sweetheart in metaphorical terms:

C'est mes deduiz, c'est mes deport. …

c'est ma main destre, c'est ma dame,

c'est moi meïsmes, car c'est m'ame.

c'est mes solaz, c'est quan que j'ai,

c'est la santé dont je garai.

It is obvious that in such a hyperbolic statement as “she is I myself (she is my very soul)” we have a quite different type of predication from “it is I.” As Foulet himself states, this c'est moi [meïsmes] is not the equivalent of ce suis je, and has no place in the development: “comment on est passé de ce suis je à c'est moi.” This 13th century example testifies only to the spread of the oblique form, generally, in non-enclitic position; in the 12th century we would probably have found c'est giê [meïsmes] or elle est gié [meïsmes]: compare the example below from Eneas (including a similar hyperbolical identification):

Cornant remandrai ge sanz toi

ne tu cornant iras sanz moi?

Dune n'ies tu gié et ge sui tu? [4943–45]

though already in the 12th century one may also find tu es moi: “Je iere tu, tu ieres moi” (cited by Foulet, via M. Humphreys, from Barlaam et Josaphat).

⁴¹ I would go so far as to say that, if a reasonable theory could be found in favor of c'est lui as the leader-construction, one might continue to accept its priority, even in the face of contradictory evidence of attestation (if the gap were not too great). For what makes Foulet's postulation of an earlier c'est moi mainly inacceptable is the untenability of his theory, at its various stages, rather than the discrepancy (of only a few years) in the dates of his examples.

⁴² Cf. (lines 6343–45):

“Ne pout mes celer sun curage

A Jasun, kar pur la grant rage

Est chaete a terre …“ (MS. A)

“… kar parla en grant rage

Et est chaete a terre …“ (MS. B)

⁴³ The corruption c'estes vus> c'est vus, instead of being rejected with exclamation-point, is accepted, by the editors, as the correct reading (for the 13th century). Their positive reason for doing so was probably inspired by metrical considerations (the original, Quant ele sav(e)ra, he co estes vus, would have one syllable too many); the negative reason is, obviously, their unfamiliarity with syntactical practice in the 13th century. Otherwise, they would have amended the line simply by discounting the final -e of ele, as they do, precisely, in the following line, El(e) s'oscirat tut a estrus.

⁴⁴ Below is a list of all the forms of the new construction which I have been able to find (I have come across no example of c'est nous), with the dates of their first attestation:

1350 c'est lui (Miracles de Nostre Dame, vii, 35).

1371–72 c'est moi (cited from M. Foulet)

1390–1400 c'est toi (Troïlus: in Nouvelles en prose du xiv^e s., p. 182)

c'est vous(“ ” “ ” “ ” “ ” p. 272)

1450 ce sont eux (Greban, Mystere de la passion, line 16717)

1456–61 c'est eux (Cent nouvelles nouvelles, p. 91)

⁴⁵ These variants, it should be noted, were at hand for all three persons. Thus there existed in OF the following parallel pronominal paradigms of identification:

ce suis je (ce suis mon) je le suis je suis celui (cil)

ce es tu tu le es tu es celui (cil)

ce est il, etc. il le est ce est celui (cil)

Of these, the paradigm with demonstrative predicative shows a discrepancy in the third person singular: the personal pronoun subject il has been replaced by ce (I have found only one example of il est celui, as againt 60 of c'est celui; here we have absolute conformity with the noun type, where the subject ce is also the norm). On the other hand, if we compare c'est celui (and c'est Jean) with il l'est and ce est il, we find an interesting study in equilibrium: in the first, the subject is neuter; in the last two, the predicative is neuter. There was evidently a feeling in OF that the same being should not be twice individualized in the same predication: if the predicate designated an individual (celui, Jean) then the subject must be abstract (ce); if the subject individualizes (as personal pronouns regularly do), then the predicate should be abstract (le, ce). One could proceed from the abstract to the concrete, or vice-versa, but, preferably, not from one concrete term to another (at least, not in the third person: in je suis celui (Jean) we have double individuation). In the noun type, there was (practically) only one construction: the subject was regularly neuter, since a noun normally individualizes (unless used adjectively: cil [il] fut grand seigneur). But, in the pronominal type, it was possible to choose, in the third person, between individualizing subject (“Il est ce” [il l'est]) and individualizing predicative (c'est celui). In general, the first was preferred; when, however, the predicative is further defined by a relative clause, the individualizing celui was usually employed: ce est il but c'est celui qui. …

⁴⁶ Thus this replacement differs from the general developments: ce>le and cil>il. In the last two, the personal pronoun has simply absorbed the demonstrative, while continuing to serve in the same function; but when lui replaces celui in the pattern c'est …, the personal pronoun has enlarged its function. But, it may be asked: if c'est celui led to c'est lui, why did not je suis (vous estes) celui lead to *je suis lui, *vous estes lui? Evidently because the language already possessed je le suis, vous l'estes, in the same meaning “(I am the one in question”), and this was considered preferable because of the neuter predicative (le), whereby double individuation was avoided.

⁴⁷ In the new construction we find, of course, the oblique case of the personal pronoun, after the model of c'est celui. But why was c'est celui preferred to c'est cil; and why had one to say “je le suis” instead of *je suis il (lui)? Why was the accusative felt more appropriate than the nominative for a predicate pronoun? One even finds in Provençal, according to Ronjat (p. 530), lis es—“[ce] les est”, and ce les est itself may be found in some French dialects. This preference for the accusative seems to be a general tendency, as is shown by the colloquial Eng. it is me (and also by similar constructions in the Scandinavian languages and others; cf. Grimm, Abhandlungen zur Litteralur und Grammatik [Berlin, 1866], p. 240). As concerns OF at least, I would assume a desire to subordinate, somewhat, the predicative: to distinguish between predicative and subject by refusing to the former the insignia of the latter. And this tendency (of “syntactical dissimilation”) should be seen together with the extension of the neuter discussed above: in line with the attitude that subject and predicative should not both have individualizing force (one of the two being sacrificed by becoming neuter), it was also felt that subject and predicative should not both be nominative, and here only the predicative could be sacrificed, by becoming “oblique.” Evidently it was intolerable that both terms of the equation “A is B” should have equal force, for (as Jespersen points out, PG, p. 154) this rarely represents a true mathematical equation.

⁴⁸ This expression is used by Foulet himself to describe the last few stages in the development of the new construction: “Et ainsi se terminera un curieux chassé-croisé: ‘c'est il’ a imposé son verbe à‘ce suis je,’ d'où ‘c'est [moi],’ mais à son tour ‘c'est moi’ va imposer la forme de son pronom à ‘c'est il,’ d'ou ‘c'est lui’” (p. 65).

⁴⁹ Or, to use the terminology of Jespersen (op. cit., p. 303; also MEG, iii, 39 ff.), c'est Jean answers an “X-question”, while ce est il answers a “nexus-question” (“is it true that…?”). The same two types of questions are known in German as Bestimmungsfrage and Bestätigungsfrage, respectively.

⁵⁰ The same function was also possible in the first and second person:

“Sire chevalier, estes vous cellui que voulez avoir le treu sur ma forteresse?”—Et cellui lui respond: “Ce suis je voirement.” Mel., 300

“Je suis Jhesus que tu meis / En ton sepulchre …”

“Sire”, dis je, “se a la place / Ou ge te posai me menoies

Et ton sepulchre me mostroies, / Donc savroie que ce es tu.“ Nic. B., 791–798

Here, ce suis je and ce es tu are concerned with the “nexus-problem” whether-or-not, exactly as was true in the examples above with ce est il (see, however, below).

⁵¹ Here, ce est il splits into two an entity first presented as a unit (le roi, son amant): il refers to the king, to the lover, only as objects of observation, as physical presences without labels; ce, in turn, refers only to the labels “king”, “lover.”

⁵² In these last examples, just the opposite of the two above, we are first offered two supposedly separate entities (the person present, and “another” person existing as a name), which are then declared to be the same. It could be said that here ce est il is really concerned not with the truth of an identification (which has been already suggested), but with the fact, itself, of identity. But this type was infrequent.

⁵³ The echo-like quality of ce est mon was first pointed out by Spitzer, op. cit., though he believed that this corroborative device was limited to characterization.

⁵⁴ Can we be sure that, in the old construction, il refers to the person-present and ce to the person-named? Theoretically, “il est ce” could mean “the one just named is the person present” as well as the reverse. However, we have only to fit ce est il into the framework of the pronominal paradigm to see that, whenever a label is involved, this is always designated by ce:

ce suis je = I who am before you am the one just named ce es tu = You who are before me are the one just named ce est il =He who is before us is the one just named

⁵⁵ One might also include here a third type, in which ce suis je is followed by a limiting relative clause:

S'il porte charge et pesant fais, / ce ne suis je pas qui le fais, / Mes Juifz

Greb., 24349–51

Here, our formula would be used to answer the question: “who is the one who … ?”—a question which is only implied, never formally expressed, in the context. The construction with limiting relative clause may be found in all three persons, and also with the noun type (c'est Jean qui …). But this periphrastic type, so common today, was comparatively rare in OF (see above, n. 17).

⁵⁶ With the exception, not only of French, but also of Provençal; cf. Ronjat, op. cit., pp. 529–530. As for Italian, it is possible that we have a blend of the old and the new constructions. In sono io, sei tu, the personal pronoun is obviously the subject; in è lui, however, lui may be interpreted either as subject (v. Tommaseo-Bellini, s.v. lui) or as predicative. Foulet evidently holds to the second interpretation, if we may judge by his rather vague words (Rom., lxii, 51): “L'italien pourrait peut-être nous aider ici [i.e., in finding parallels for the development he has postulated for OF], car il semble bien qu‘à l‘époque moderne, et depuis plus longtemps peut-être dans certains cas [?], des innovations s'y fassent jour qui ressemblent fort a quelques-unes de celles qui se sont produites en français au cours de l'histoire du pronom personnel.” If è lui does, indeed, represent the new construction, I suggest that its origin is to be found in the demonstrative type: è colui—exactly parallel with the OF development. Compare, for example: “Tu dicevi che eri colui il quale questa notte avevi ucciso 1'uomo?”: Boccaccio (cited by Tommaseo-Bellini).

⁵⁷ It might be said from another point of view, that the question “who is that?” is not only paradoxical in the situation just described, but is a paradoxical predication in itself, whenever used. For who refers to a person, and that to a thing (“what person is that thing?”). One will, of course, find nothing parallel in Latin, where only quis est (hic) and quid est (hoc) were possible.

⁵⁸ Originally, the concept “that” or “it” did not exist even for a third person: little wonder, then, that the speaker should hesitate to think of himself as an abstraction.

⁵⁹ Indeed, “I” is that which is presupposed by the act of speaking; it is always an “I” that speaks. The inherent difficulty of thinking of oneself “object-ively” is suggested by the fact (mentioned by Grimm, op. cit., p. 240) that, in Indo-European, “I” has no accusative case, the form me representing a different stem.

⁶⁰ This means, of course, that “I”, in such a situation, is used with special significance. Ordinarily, even in self-identification, this pronoun means only “I whom you see before you, I who address you”; this is true when ce suis je is used to answer the question “Estes vous … ?”, and also when a name has first been mentioned (“[nous querons] Jhesus …”

—“Ce suis je”). But in the sequence “Qui est ce [que je vois]?”—“Ce suis je”, such a relativistic reference would amount to nonsense: ce suis je=*“I whom you see before you am the one you see.” Here, “I” must mean: “I whom you know as an individual.”

⁶¹ In most cases, it is true, je suis Jean will be found in answer to the question “qui estes vous?” In such a case we have, obviously, perfect logical sequence.

⁶² In general, this borrowing of the third-person formula by the speaker for application to himself, was limited to cases in which the one who identifies himself is unseen or imperfectly seen by his audience, and, consequently, is forced to realize that he exists for them only as a sound or shape:

[Li garçons:] Sire, vous n'alés mie bien: / vous querrés ja en cest celier.

[Li aveugles:] A! mere Dieu … ! / Ki es ce qui si bien m'avoie?

[Li garçons:] Preudons, se Jhesus me doint joie, / çou est uns povres triquemers.

G+A, 19–24

Quar mult li vint a grant merveille / Quant il senti lez lui la dame:

“Qui est ceci?”—“C'est vostre fame.”

Ruteb., ii, 136

have found one example, however, in which the speaker is in full sight of his audience:

[La Vierge:] Qui es tu, qui vas par ci?

[Théophile:] Ha, Dame, aiez de moi merci! / C'est le chélis

Théophile, li entrepris, / Que maufé ont loié et pris.

Ruteb., ii, 257

Indeed, here, the third-person formula is used to answer a question in the second person; it is “je suis …” which would have been, actually, the more logical response. The motivation must be that of humility. “Je suis …” often had a nuance of bold self-confidence (“Know who I am!”) And “li chétis Théophile” wished to present himself in the most modest way possible to the Virgin—seeing himself as he must appear in her eyes: a phenomenon (which has disturbed her).

⁶³ In the question “estes vous ce?” the personal pronoun has the same limited, personalized reference mentioned above in reference to ce suis je; vous must mean not “you whom I see before me, you whom I address” but “you whom I know as an individual, and think to recognize.” Here, “you” is as truly a proper name as “I.”

⁶⁴ As concerns the priority of “I” over “you”, I may quote the following words of Grimm (op. cit., p. 238): “Unverkennbar sind die beiden ersten leiblich gegenwärtigen personen viel lebhafter als die dritte entferntere, die erste musz aber für die innerlichste, die zweite fur die vertrauteste gelten. alle rede hebt an mit dem was die erste person denkt, will, sagt; sie ist die erkennende, mittheilende, bittende, fragende, lernende, die zweite, der sich jene aufgeschlossen hat, die theilnehmende, kundige, erbetene, antwortende, lehrende. das ich ist ein über die lippe des redenden an das ohr des du schallendes wort, und begehrt gehör, billigung oder antwort, die zu gewähren von der zweiten person abhängt, welche ich darum die trauliche oder gewichtige nenne.”

⁶⁵ From the middle of the 14th century until the first part of the 16th, I have found only 11 examples of c'est lui as against 34 of ce est il (1 to 3); during the same period, however, I found 15 examples of the new construction c'est moi as against only 7 of ce suis je (2 to 1).

⁶⁶ That the extension from the third person to the first and second does not represent an inevitable development, of a mechanical nature, is shown by the situation in Italian, where the new construction è lui (if this is indeed equivalent to c'est lui, which seems most likely) has had absolutely no influence on the traditional forms sono io and sei tu. It it true that Italian did not possess the machinery to facilitate this development quite so smoothly as was the case in Middle French: there were no pronouns of the first and second person to correspond to moi and toi, so that è lui could have led only to *e io, *è tu—representing, perhaps, an awkward juxtaposition. But the juxtaposition [it] is I, is you was also awkward in English at the beginning (see below, Appendix)—and nevertheless became accepted.

⁶⁷ He does, it is true, analyze the first attestation of c'est lui—but only to reject it precisely because of its individual flavor, its climactic force.

⁶⁸ I was able to find only 2 other examples of c'est vous in the 15th and early 16th centuries, as against 15 of c'estes vous.

⁶⁹ In this connection, it is interesting to compare the example of est ce toi? cited above from Troïlus, with the following example of es tu ce? from Greban, fifty years later, used in a context superficially identical with the one just discussed. Here, too, we have a grieving parent gazing upon the altered bodily appearance of a beloved son, and uttering the same distraught question: “can this be you?” But the emotion of the speaker lies too deep to be expressed by the intellectualized est ce toi?

Filz beneuré, filz bien assaissonné,

fils gracieux, filz bien moriginé,

corps des vivans mieulx complectionné,

o beauté pure,

Choix des humains, fleur de toute nature,

riche couleur, parfaicte pourtraiture,

regard piteux, tres benigne stature,

face sacree,

Face luysant, franchement figuree,

es tu ce la, si tres desfiguree,

blesme des yeux, tout de sang pourpree,

dur entremes?

Filz, es tu ce, je ne te congnoy mes,

qui vas mourant? ne t'adviengne jamès;

en croix te rendz, a la mort te submès

sans desservir.

In this lament of Mary over Christ on the cross, the primitive atmosphere reappears. No distantiation is possible between this mother and her son, bleeding and disfigured though He is. And, though in her grief she must cry out to Him: “je ne te congnoy mes”, she knows Him, with an eternal knowledge; and her query “est tu ce …?” begins with His reality, and questions only appearance (“ce, qui vas mourant”).

⁷⁰ There seems to be a necessary connection between the pronominal type c'est moi and the noun type c'est Rollant: in English, where we find it is I [me], it is also possible for a speaker to introduce himself by “this is, it is so and so”; such a formula goes back at least to the 14th century:

Grewosly up starte hee / And sayd: “What devill art thou?” …

“Mercy” she cryed nowe… / And sertes, it is Imayne,

That is comyn to you …“

(I pomedon, 7195)

Contrariwise, in German, Spanish and Italian, where the old pronominal construction “I am [that]” still prevails, it is, I believe, impossible to introduce oneself by name in the third person.

⁷¹ In two volumes of Larivey, I was able to find not a single example of the old construction—and dozens of the new.

⁷² Cf. the article of Leo Spitzer, SP, xxxvii (1940), 565–584.

⁷³ Studies in English Syntax (1906), ch. iii: “The position of Words as a Factor in Syntax”, pp. 77–86. In this chapter, the formulas of identification represent only one of several types which, according to Smith, have become altered in meaning because their word-order led to reinterpretation. But the alteration with which he is concerned, in the case of our type, is not the reversal of subject and predicate, but simply the use of the accusative for the predicative: he is not primarily interested in the development leading up to it is I, in which it has become the subject: he takes this development for granted, and stresses only the fact that once it is I arose, the predicative pronoun, finding itself in “objective territory”, came to be felt as object (> it is me).

⁷⁴ This coincidence did not exist in OF: the “Paradigm ii” which Foulet (partially) postulated for OF, which should contain c'est il with ce subject (*c'est je, *c'est tu, c'est il) was not convincingly attested (see above, note 34).

⁷⁵ I may add that Smith's chronology is also untrustworthy in regard to the type in which he was primarily interested: that with accusative predicative. His first (unquestionable) example is from Marlowe; but I have found the construction “art thou hym?” in the 14th century (Sir Perc. of Galles, 2041), and “I him am” in the 13th (Southern Passion, 1998).

⁷⁶ The first examples attested by Smith are from Chaucer (if we exclude certain examples of the noun type “it are my friends” which he has mistakenly included here). And I myself have gone through nine volumes of 14th-century texts before Chaucer without being able to find a single example of it am I, it art thou, it are ye.

⁷⁷ In the Northern Passion, the question asked at the Last Supper by Judas and the other disciples: “Numquid ego sum?” (“am I the one who will betray thee?”) is rendered, lines 258 and 268, by “Was it 0uзte I?”; here, for both lines, all 8 MSS used by the editors agree (except that in one case the Latin original is found). We may also note, line 532, the reading of one MS: “I it is” (the other 7 showing I it am or I am he). The new construction is also found with the second person in this text: in line 270, 7 MSS read “thou it is”, and one “thiselve it is” (most of them give the curious alternation: “þou it arte and þou it is” = “thou art that person, and that person is thou”). Finally, in 1. 703, one MS reads “þis ilke ys þow þat …”, though the majority have “þou art one þat …”.

⁷⁸ The one exception is an example from Pol. Poems, Rolls, ii, 57 (borrowed from the NED and there dated 1401): “It ar зe that stonden bifore.”

⁷⁹ Foulet cites this same example, but his attempts to explain it away reveal a lack of familiarity with negation in English. He seems to believe that Eng. not = Fr. ne, whereas, obviously, not (ne … ought) is a reinforced negative, an emphatic negative exactly parallel to Fr. ne … pas, point—which, likewise, cannot be followed by the subject.

⁸⁰ The emphatic word-order I it is, thou it is is evidently based on that of the noun type: the arrangement “John it is” was very common in ME (King Horn, 313: “He wende that Horn hit were”). Since the normal word-order of the new pronominal type follows that of the noun type (it is John, it is I), it is understandable that this pronominal type should also borrow the emphatic order of the latter—and, indeed, because of its “newness”, borrow this first of all.

⁸¹ It is, of course, in general, a very dangerous procedure, in dealing with a problematical construction, to disregard verbal agreement as a clue to the subject, and to postulate “anticipation.” Indeed, it may prove a double-edged sword: for instance, in the examples above of thou [I] it is, which I have interpreted as containing thou-predicative in emphatic position, I have, obviously, chosen to accept the evidence of the form is as proof that it is the subject; theoretically, one might argue that we have here the normal pronominal word-order of Paradigm 1 (ic hit eom, pu it eart), with thou subject, it predicative, and is in agreement with this predicative. But, here, I would counter that agreement with the predicative is conceivable only when the predicative somehow “outweighs” the subject. If we find regularly in the early language the type “it are my friends” (corresponding to French and German practice), it is because the plural, naturally enough, may be felt to outweigh the singular. It is also easy to understand that the first and second persons may be felt to have a greater emotional weight than the third person—and, therefore, be anticipated by the verb, as in the occasional it am I, it are ye (cf. also the established construction of contagion: “it is I who am … ”). But, for that very reason, we can hardly imagine the predication *thou is it: a third-person predicative could never “outweigh” the subject thou.

⁸² That Smith, who has worked on English word-order, could postulate the word-order C-V-S for it am I is difficult to comprehend; that this was accepted by Foulet is no doubt due to the fact that the French scholar is thinking in terms of ce suis je, where we do, indeed, find the word-order C-V-S. But Foulet has completely overlooked the fact that the complement ce is a demonstrative, while the complement it is a personal pronoun; and the two types observe a different word-order, in both languages. It should be obvious that the real parallel to the OE paradigm with hit is the OF paradigm with le: ic hit eom—je le suis correspond perfectly. Foulet, however, has consistently ignored the type je le suis, and thinks of the construction with ce as the unique formula of identification in French. Accordingly, he can see the English formula only in connection with ce suis je: when he finds in OE, ic hit eom, this represents to him a “different” word-order from the French; when he finds, in the 14th century, it am I, this means that English has adopted the “same” word-order as French (“Impossible de trouver entre deux langues une ressemblance plus complète de structure et de procédés”: p. 55) ! He does not realize that it am I, as he interprets it, could correspond only to (the impossible) *le suis je.

⁸³ Thus Smith's “Paradigm ii” is completely mythical, just as was true of the “Paradigm ii” which Foulet (partially) postulated: *c'est je, etc. Both scholars have constructed, out of thin air, transitional paradigms, built to last only a few decades, in order to bridge over gaps in their theories. But genuine paradigms (those formed by the language itself) have deep roots, and are slow a-building and slow to die.

⁸⁴ If the new word-order were alone involved, we would simply have found: I am it (as Foulet himself points out). Incidentally, this form does appear (earlier than it is I): I have found an example from the Southern Passion (late 13th century): “‘Mayster,’ quaÐ Iudas þo ‘y nam hit nouзt ywis’” (824).

⁸⁵ Skeat, The Four Gospels (in Anglo-Saxon etc.), 1887; I quote from the Corpus MS. It may be noted that, in the examples with se predicative, the demonstrative serves as antecedent of a relative clause. In OF, too, c'est celui was mainly followed by qui. …

⁸⁸ I must confess that I have not, in my reading, come across hyt is se in OE: only þaet is se, þis is se. But given the easy alternation of the three neuter pronouns, it is perhaps legitimate to postulate this particular form. (Or else, one might assume the development: þaet [þis] is se>that [this] is he>it is he.)

⁸⁷ It is interesting that in my first example of the third person formula we find the emphatic negative arrangement that shows clearly the predicative function of [s]he: “it was not shee.” This is clear proof, if any were needed, that the supposedly “ambiguous” it is he has had only one interpretation from the beginning. This means, too, that Smith's chronology is even more out of joint than was suggested above: “Paradigm iii” goes back (at least) to the middle of the 13th century (instead of the 15th).

⁸⁸ It might be said that, in the word-order I am it cited above (note 84) from the Southern Passion, we have the influence of this new type I am he; and that, contrariwise, in I him am (note 75), from the same text, the new / am he has been drawn into the orbit of the traditional I it am! (And perhaps the accusative form him, which here replaces it, shows that this neuter itself was felt as an accusative [= OF “je le suis”].)

⁸⁹ A paradigm corresponding to OF je suis celui, lu es celui, c'est celui. I have not attested ic eom se, but I have several times found this pattern represented in the second person: þu eart se: “Eart þu se Ðe toweard is, oþþe we oÐres andbidian sceolon?” (Aelfric Hom., ed. Thorpe, i, 480: cf. also Alfred the Great, Psalms, iv. 1).