¹ The term is defined below. See also Mitchell, B., Old English Syntax, 2 vols. (Oxford, 1985), §1937 (hereafter OES).CrossRefGoogle Scholar

² Gorrell, J.H., ‘Indirect Discourse in Anglo-Saxon’, PMLA 10 (1895), 342–485, at 458.CrossRefGoogle Scholar

³ Old English Modal Verbs: a Syntactical Study, Anglistica 26 (Copenhagen, 1990)Google Scholar, Introduction and §6.1.1.

⁴ ibid. §§2.2 and 3.2. My modal verbs in this book, as in the present study, include nine words: agan, cunnan, *durran, magan, *motan, *sculan, purfan, willan and wuton.

⁵ ibid. §§5.1.1 and 6.1.1. A list of Old English volitional expressions (as I define them) is given in the introduction to Part II of the same book.

⁶ In this table, as in other relevant ones in this study, MV and IF stand respectively for modal verbs and inflexional forms of simple verbs. For Ælfric's prose, I have examined all his non-liturgical narrative pieces as set out in Clemoes, P., ‘The Chronology of Ælfric's Works’, The Anglo-Saxons: Studies in Some Aspects of Their History and Culture Presented to Bruce Dickins, ed. Clemoes, P. (London, 1959), pp. 212–47, esp. 218.Google Scholar

⁷ For the dominance of the pret. pl. -en ending in dependent desires in Orosius (ed. Sweet; see n. 15), see my article ‘Modal Verbs in Noun Clauses after Volitional Expressions in the Old English Orosius’, Stud, in Eng. Lit. (Tokyo), English Number 1979, pp. 115–37, esp. 117. For later usage, I find, for example, some fifty forms ending in -on/-an but only one in -en in the corresponding contexts in Ælfric's Lives of Saints, ed. W. W. Skeat, EETS os 76, 82, 94 and 114 (London, 1881–1900). For ambiguous preterite plural forms in general, see Campbell, A., Old English Grammar (Oxford, 1959), §735(g).Google Scholar

⁸ Standop, E., Syntax und Semantik der modalen Hilfsverben im Altenglischen: Magan, Motan, Sculan, Willan, Beiträge zur englischen Philologie 38 (Bochum-Langendreer, 1957), 45Google Scholar: ‘What is more, there still remains in Old English a distinct preference for a mode of expression which, to our linguistic instinct, works pleonastically and tends to ‘overdefine’ syntactic functions. Indeed, in the sentence type ‘Ich befehle, daß du gehen sollst [I command, that you shall go]’, desired relationship is doubly expressed: through ‘befehlen [to command]’ and ‘sollen [shall]’. Modern language replaces this by ‘Ich befehle dir zu gehen [I command you to go]’, and this is so in German as well as in English. However, Old English goes even a step further in that it uses this self-same sentence type for principal and subordinate clauses of which the subjects are identical: ‘ich beabsichtige, daß ich … will [I intend that I will…]’; ‘ich bitte, daß ich … darf [I pray that I may …]’, and so on. The infinitive, though in use already in Old English, is not so characteristic as are subordinate constructions. These are another indication of the well-known fact that sentence construction in Old English still remains very close to parataxis.'

⁹ OE Modal Verbs, Tables 22, 29–31 and 38–40.

¹⁰ See, e.g., the quotation from Standop on p. 83.

¹¹ Syntax und Semantik, p. 169.

¹² We could have included ^*motan and (ne) þurfan in what may be called the variant of the ‘identical-subject type’ in which one asks for permission on behalf of somebody else; the percentage of ‘paratactic’ uses is then 59.1 per cent in early prose (with six such examples) and 56.5 per cent in late prose (with seven examples).

¹³ Ogawa, , ‘Modal Verbs in Orosius’, p. 133Google Scholar. Wisher and Doer in the quotation refer to the two key agents in desired relationship.

¹⁴ Some more examples of this kind of negation of modal verb, as opposed to negation of main verb as in ModE should not, must not, occur in Bo 18.26, ÆCHom II (ed. Godden) 37.120, WHom 13.34; cf. Afar; 20.17, WHom 4.25 (for these abbreviations, see below, n. 15). See also Standop, , Syntax und Semantik, p. 104Google Scholar. Restriction of þurfan with ne to early Old English is also seen in the ‘non-identical-subject type’ of requesting. The total absence of þurfan in dependent desires strikingly characterizes the late usage.

¹⁵ Citations from the main texts in this study are made from the following editions: The Old English Version of Bede's Ecclesiastical History of the English People, ed. Miller, T., EETS os 95 and 96 (London, 1890–1981)Google Scholar; Bischof Warferths von Worcester Übersetzung der Dialoge Gregors des Grossen, ed. Hecht, H., Bibliothek der angelsächsischen Prosa 5 (Leipzig, 1900 and Hamburg, 1907)Google Scholar; King Alfred's West-Saxon Version of Gregory's Pastoral Care, ed. Sweet, H., EETS os 45 and 50 (London, 1871)Google Scholar; King Alfred's Orosius, ed. Sweet, H., EETS os 79 (London, 1883)Google Scholar; The Old English Version of the Heptateuch, ed. Crawford, S.J., EETS os 160 (London, 1922)Google Scholar; and Two of the Saxon Chronicles Parallel, ed. Plummer, C. (Oxford, 1892–1899)Google Scholar. For Orosius, the edition by Sweet has been cited because the detailed statistics were originally compiled from his edition, but the examples and figures are not significantly affected if the more recent edition by Bately, J. (The Old English Orosius, EETS ss 6 (London, 1980))Google Scholar is used instead. The texts are referred to by the short titles set out in Mitchell, B., Ball, C. and Cameron, A., ‘Short Titles of Old English Texts’, ASE 4 (1975), 207–21Google Scholar. References are to chapter and verse of Ælfric's translation of Joshua but are otherwise to page and line (followed, in the case of the Chronicle, by annal in parentheses).

¹⁶ See below, p. 90, n. 20.

¹⁷ For OE man as a substitute for the passive construction in general, see Mitchell, OES, §§369 and 747; Fröhlich, J., Der Indefinite Agens im Altengliscben (Bern, 1951), throughout.Google Scholar

¹⁸ It is thus an example of what Fröhlich calls man5, i.e. man is ‘definierbar’ (Der Indefinite Agens, p. 102, and throughout)Google Scholar. By contrast, man in Or 264.8 and 226.15 (to be discussed later) are classified by him as man3 {man is ‘allgemein’) and man4a {man is ‘speziell’ but ‘unbekannt und nicht erschließbar’), respectively (ibid. pp. 51 and 78).

¹⁹ See Frary, L. G., Studies in the Syntax of the Old English Passive with Special Reference to the Use of Wesan and Weorðan, Language Dissertations 5 (Baltimore, MD, 1929), 22.Google Scholar

²⁰ For details see Ogawa, OE Modal Verbs, §§5.5.3 and 5.6.3. Willan to ask for favour in dependent requests (OED s.v.v. 6b, 26b) is not attested in Old English except in two sentences in late prose; on this see my ‘OE ^*Sculan/Willan in Dependent Requests: a Note’, Philologia Anglica: Essays Presented to Professor Yoshio Terasawa on the Occasion of His Sixtieth Birthday, ed. Oshitari, K.et al. (Tokyo, 1988), pp. 53–9.Google Scholar

²¹ However, this absence of ^*motan with a passive infinitive does not seem to mean that the resultant voice of a verb determines the choice between the ‘identical-subject type’ of clause with a ‘definite’ subject and ^*motan on one hand and the non-identical-subject construction with man and a simple verb on the other, with the latter to be used as an ‘automatic substitute’ for the former when this would result in the ‘unidiomatic’ passive infinitive with the modal verb. Basic paradigms of Old English verbs do allow such a form, though examples are more common in works of translation; see Mitchell, OES, §§751–2; M. Callaway, Jr, The Infinitive in Anglo-Saxon (Washington, DC, 1913), pp. 83–8. For examples, see (dependent desires) El 686 ‘Ic þæt geswerige þurh sunu meotodes, / … þæt ðu hungre scealt / for cneomagum cwylmed weorðan’; Bede 374.2 ‘he … wæs ða broðor biddende, þæt he ðær… bebyrged beon moste’, and (other clauses) Or 128.5 ‘þa Darius geseah ðæt he oferwunnen beon wolde’; and ÆCHom I, 282.19 ‘Fæder, and Sunu, and Halig Gast ne magon beon togædere genamode’; we also find a simple passive verb in the ‘identical-subject type’ in Chron A 8.19 (167) ‘bæd þæt he wære Cristen gedon’. As we shall see, choice of the man construction is very much a matter of style.

²² An early Old English example of man wolde in subordinate clauses occurs in GD 219.11 ‘þa gesawon hi unfeor þanon ænne ofen inæledne, se wæs gegearwod to þon þæt man wolde bacan’. Here, however, man is, as Mitchell (OES, §2978) points out, identical with the ‘logical subject of the preceding adjective clause’ which is understood. See also LS 10 (Guthlac) 18.4.

²³ The last part of the material thus belongs in fact to the Middle English period. But for practical purposes it is treated collectively as late Old English. For the character of the language in the Copied Annals, see Clark, C., The Peterborough Chronicle 1070–1154, 2nd ed. (Oxford, 1970), pp. xli–xlv.Google Scholar

²⁴ Man in this sentence is rightly classified by Fröhlich as man4a (see n. 17). It distinguishes itself from the use as ‘quasi-identical subject’ evidenced in Josb. IX.15 and the Chronicle (see below) in that it is not the senators who do the building. Gorrell, (‘Indirect Discourse’, p. 406)Google Scholar appears to mistake the sentence for an example of hit gewearð ‘it happened’.

²⁵ There is another example of willan in a remotely related clause dependent on unrædes (Chron E. 141.18 (1011)). But this is not included in the figures of Table 5.

²⁶ One is reminded here of expressions common in laws such as Law Atr I Prol ‘Ðis is seo gerædnys, ðe Æþelred cyning 7 his witan geræddon.’ But there seems to be little need to assume that the use in the Chronicle of man after such expressions of royal agreement could not have developed without the basis of a legal idiom; the laws contain only one example of the construction being discussed (Forf2), as far as we can tell from Liebermann's Wörterbuch, F. (Die Gesetze der Angelsachsen, 3 vols. (Halle, 1903–1916) II, s.v.gerædan)Google Scholar. For the use of the word ræd in laws and chronicles, see Liebermann, F., The National Assembly in the Anglo-Saxon Period (Halle, 1913), p. 12.Google Scholar

²⁷ ‘Indirect Discourse’, p. 367. Gorrell gives a wrong reference to his truncated version of the example (‘se cyning gerædde þæt man sceolde habban gemot’) as ‘Chr., 250, C, 20’. What he means – Chron C 250.20 (ed. Thorpe) – is parallel to our Chron E. 133.31 discussed below.

²⁸ See, e.g. Whitelock, D. et al. , The Anglo-Saxon Chronicle: a Revised Translation (London, 1961), pp. 120 and 118 respectively.Google Scholar

²⁹ ibid. p. 83. As this example shows, the ‘quasi-identical-subject type’ of dependent desires need not be supported by the use of a modal verb. See also Chron E 145.1 (1014).

³⁰ The Anglo-Saxon Chronicle, trans. Garmonsway, G. N., 2nd ed. (London, 1954), pp. 133 and 137Google Scholar. He extends the interpretation to Chron E 133.3. But I should prefer not to take it as of the ‘quasi-identical-subject type’; see above, p. 93.

³¹ The National Assembly, p. 12. The reference is given as a footnote.

³² The third clause in Chron E 177.20 (1052) 7 gerædde man þa þ þa scipu gewendan eft ongean to Lundene. 7 sceolde man setton oðre eorlas 7 oðre hasæton to þam scipum’ could be taken as one such case. But, on the grounds of its inverted word order (which is parallel to the initial ‘7 gerædde man’) and the contrast to a simple verb in the preceding noun clause, I should prefer to take it as an independent clause.

³³ The C version has woldon here.

³⁴ Comparison in terms of the number of modal verbs as against that of simple verb forms within each type does not prove as decisive (see the table below). But the decline of modal verbs in the ‘identical-subject type’ is at any rate clear enough.

³⁵ See, e.g. Rübens, G., Parataxe und Hypotaxe in dem ältesten Teil der Sacbsenchronik, Studien zur englischen Philologie 56 (Halle, 1915)Google Scholar; Clark, C., ‘The Narrative Mode of The Anglo-Saxon Chronicle before the Conquest’, England before the Conquest: Studies in Primary Sources Presented to Dorothy Whitelock, ed. Clemoes, P. and Hughes, K. (Cambridge, 1971), pp. 215–35.Google Scholar

³⁶ Syntax und Semantik, p. 169: ‘The interaction [between modal verbs in dependent clauses and introductory verbs in principal clauses] in the reverse direction consists in the fact that the respective use of modal verbs is a contributory determinant of the sense of the principal verbs, particularly those of informing, perceiving and desiring.’

³⁷ Wilde, H.-O., ‘Aufforderung, Wunsch und Möglichkeit’, Anglia 63 (1939), 209–391 and 64 (1940)Google Scholar, 10–105 (the quotation occurs on p. 67 of the latter volume): ‘The object clauses and subject clauses with modal periphrasis … represent a special, so–called divergent type. The more strongly the preceding clause expresses neutrality and eventuality, the greater and more necessary the use of periphrasis becomes.’ An aspect of the decreasing use of the ‘identicalsubject type’ of noun clause in favour of the infinitive construction in Middle English is discussed in my article ‘The Periphrastic Subjunctive in ME were/had lever Constructions’, Stud, in Eng. Lit. 49 (Tokyo, 1972), 55–66 [in Japanese].Google Scholar