idea: S. Mar. 10. 65 And þeh hit eall gelumpe þæt ænig hæfde pa geðincoe þæt heo mihte mæden beon and eac cildes moder, pehhwedere nefullcumð næfre nan to þære mærde, þæt heo pone ilca sune gebere odde ægne his gelica. одбе

NOTE. In Middle and Modern English, all, of course, has its separate history as a concessive connective. See NED., article albeit, and cf. the following: Hal. Meid. 43. 23 Ne telle þu nawt eðelich, al beo pu meiden, to widewen ne to iweddede; Leigh Hunt, Autobiog. (London, 1870) p. 369 I could have got it, had I been wise, for a third part of the sum, albeit it was neatly bound.

Thus all becomes, in the Middle English period, virtually a concessive conjunction. This originates as an adverb within the concessive clause, as the examples given above make clear; and the history of the particle is not parallel to that indicated by Mensing for MLG., MHG. al and alein (Mensing, pp. 60–69).

A mannerism consisting in the use of nu after both Jeah and other conjunctions-rather as a mere expletive than as an intensive-is characteristic of the Old English Boethius. This nu has usually no definite meaning, and no influence on the construction of the clause. I cite a few out of numerous examples: Bo. 27. 22 peah hy nu ece wæron; 44. 4; 46.30; 48. 17; 69. 1; 72. 30; 101. 8. The same expletive in deahclauses is found in a few passages of Cura Pastoralis : 41.2; 101. 11; 265. 6

The deah-clause is sometimes periphrastically introduced; thus BH. 124. 14 has peah pe pat wære pot for the quamvis of the original. The periphrasis consists in the use of the copula or of a colorless verb meaning 'to happen,' 'to befall,' followed by dat. In some cases, the intention is, apparently, to emphasize the

remoteness or deplorableness of the idea contained in the concessive clause: S. Mar. 10. 65 (see p. 21 above); ÆH. 1. 242. 21 Ge sceolon beon geornfulle to eower agenre dearfe, þeah hit swa getimige þæt se lareow gimeleas beo; PPs. 4. 5 þeah hit gebyrige þæt ge on woh yrsien, ne scule ge hit no þy hrapor þurhteon, þe læs ge syngien, where the Vulgate has simply: Irascimini, et nolite peccare; probably Inst. 477.21 (see p. 19 above). In Dial. 34. 2 there is probably an effort to render fortasse: þeh þe hit gelumpe þæt him hwilc man þe hine ne cupe ongen come (Si quis illum fortasse nesciret, salutatus). In other cases the construction serves rather to give weight and impressiveness to the statement of the deah-clause. Such is its value in the passage from Bede quoted above (BH. 124. 14); Dial. 13. 1 þeah þe heom gelumpe, þæt hi utan on heora lichaman mænniscra lareowa lare wana wæron, þehhwæðre heom gelamp, þæt hi mid haliges gastes gife in heora heortan ingehigdum onbryrde 7 gelærde wæron; Wulf. 78. 17 and deah þæt sy, þæt fela manna Antecrist sylfne næfre his eagum ne geseo; 79. 19 and deah þæt geweorde þæt ure ænig, þe nu leofað, þonne ne libbe; 227.8 and þeh hwam gebyrige, þæt his fyr ut gewite, nis þæt alyfed to begetenne. The last case, however, is to be interpreted as periphrasis in the making, not developed as a separate construction, since gebyrige may be regarded as retaining its full meaning. In the other passages quoted, where the periphrasis is plain, its effect is one of heaviness and clumsiness; but, by acting as a sort of retard, it undoubtedly serves to fasten attention upon the concessive clause.

C. The Mode of the Clause.

The problem of modes in concessive clauses cannot be investigated in the mass; each connective requires special study. By far the greatest importance in this regard belongs to deah, by which simple concessive clauses are ordinarily introduced. An index of deahclauses would show the overwhelming preponderance of the optative with this conjunction. But a word must be said as to the method to be applied in such investigation.

The only way to secure trustworthy results in the tabulation of modes is to make due allowance for ambiguous forms. Among these must be reckoned, in Old English, not only the preterit singular first and third person of weak verbs, but also preterit plurals in -on, except in the few texts where the distinction between optative and indicative preterit is consistently marked. It is only when unmistakably indicative forms occur that we can assert: 'In these cases, deah is followed by the indicative.' This fact has not always been duly recognized, however, by writers upon Old English. Selecting, for example, the deah-clauses from the variety of concessive clauses which Wülfing cites (2.148) as containing an indicative verb, we find that most of his examples (leaving out of account one optative: ongitan) are weak preterits in the third person singular, or plural preterits ending in -on. Only one example is certainly indicative, a passage from Boethius quoted, apparently, from Cardale's edition. For the distinction between indicative and optative plural preterit is by no means

1 An error, probably, of the late Bodleian MS., for Sedgefield, printing the end of the word from the Cotton MS., has the optative: Bo. 50. 21 þeah þu nu hwene ær sæde.

always to be found in the Alfredian texts. On this point, see Krawutschke, p. 45; Fleischhauer, p. 24.

A rough and ready way of learning the preference of the language as to the mode with deah, and a preparation for determining whether the usage was uniform for all senses of the conjunction, is to note the mode after deah-clauses of any sort-concessive, conditional, or quasi-interrogative-remembering, however, that in most cases deah is concessive. The following table, including a selection of representative texts, gives a survey of the modes with this conjunction.

The single indicative in Boethius, Bo. 31. 10, occurs in a passage supplied from the later manuscript. De Temporibus, though a brief text, has purposely been included, because its proportion of indicatives is large. One indicative in the Lives of the Saints, LS. 1. 150. 35, is synd, the same as the verb of the principal clause, and may be due to confusion or attraction.

1 Homily 27 has been excluded, as being a duplicate of a text in Rood. See LS. 2. xi.

The strongest statement, then, that we can make is that the indicative may appear after deah, but exceptionally in both earlier and later texts. These results agree substantially with those of Nader for Beowulf.1 In view of the almost negligible number of undoubted indicatives, discussion as to whether the mode after deah varies with its occasional conditional or interrogative use, or with a particular form of concession, becomes idle. We may, however, apply the usual categories of fact and supposition, to see whether these optative deah-clauses are restricted to one meaning or the other. For this purpose, we may select a work of Elfric, who rather inclines to the use of the indicative in various constructions which may have the optative (see p. 62 below, for an example of this). Of the 104 deah-clauses in the Lives of the Saints, all are concessive. Of these, 66 seem to me undoubtedly concessions of fact; the others hypothetical. And of these 66 concessions of fact, 44 have plainly optative verbs. Old English did not mark by mode the distinction between fact and supposition introduced by deah.

NOTE. In the Heliand, Behaghel finds the subjunctive practically constant after thoh (Modi, p. 46), though he mentions one or two exceptions. As for the cognate particle in Otfrid, Erdmann says explicitly (Otf., p. 90): 'Ohne Ausnahme steht... der Conj. in dem durch thoh eingeleiteten concessiven Nebensatze sowohl wo dieser dem Hauptsatze vorangeht, als wo er nachfolgt, und nicht bloss bei allgemein möglichen, sondern auch bei entschieden als wirklich und tatsächlich gedachten Ereignissen.' The usage of these works, then, agrees with Old English.

...

The mode of the swa-clause, on the contrary, is indicative in all cases where the mode can be deter1 Anglia 11. 452.