16. Ghirshman (1954: 229-30).

17. Mango (1980: ch. 1).

18. Plutarch, Mark Antony, xxvii.

19. Cambridge Ancient History, vol. vii.12, p. 180.

20. Drew-Bear et al. (1999).

21. Strabo, iv.1.5.

22. Plautus, Epidicus, iii.3.29.

23. Polybius, Histories, iii.59.

24. Vergil, Aeneid, vi.847-53.

25. pergraecari est epulis et potationibus inservire: in the dictionary of s.e.xtus Pomponius Festus of the late second century AD. The word is common in Plautus, the great adapter of Greek plays for Roman audiences in the second century BC.

26. Sawyer (1999: 37).

27. ibid.: 35.

28. The source is an Athenian sophist, Philostratus, whose Life of Apollonius of Tyana was commissioned at the end of the second century AD by the wife of the Roman emperor Septimius Severus. This is a work of devotional literature, and so its accuracy has been questioned; but Woodc.o.c.k (1966: 130) argues that archaeology shows the author was in fact well informed about details of this land so remote from contemporary Rome and the Mediterranean.

29. Wiesehofer (2001: 122).

30. ibid.: 155.

31. Itinerarium Aetheriae (ed. H. Petre, Paris, 1948), xlvii.3-4 (quoted in Mango 1980: 19).

32. Mango (1980: 25).

33. De Thematibus, Introduction, Pertusi edn, 1952, quoted in Horrocks (1997:150).

34. Procopius, Secret History, xviii.20-21.

35. Third Part of the Ecclesiastical History of John Bishop of Ephesus, trans. R. Payne Smith. Oxford, 1860, pp. 423-4 (quoted in Mango 1980: 24).

36. P. Lemerle, La Chronique improprement dite de Monemvasie, in Revue des etudes byzantines, xxi (1963), pp. 9-10 (quoted in Mango 1980: 24). The Kafirs were perhaps Muslim converts; the Thracesians were not Thracians, but from the Thracesian theme, in the west of Anatolia.

37. Leo VI, Tactica, in Patrologia Graeca, ed. J. P. Migne, cvii, 969A (quoted in Mango 1980: 28).

7 Contesting Europe: Celt, Roman, German and Slav

1. Herodotus, ii.33, iv.49. The Cynetes, aka Cynesians, may have been correctly placed just beyond the Pillars of Hercules, since Strabo, iii.1.4, calls this area, the modern Algarve, Cuneus-though he thought that it was named in Latin after its wedge-like shape.

2. Jacoby (1923: no. 70, fr. 30).

3. Strabo, vii.3.8; Arrian, i.4.6-8.

4. Tain Bo Cuailnge (Book of Leinster, 2nd Recension), 11. 4733-6, trans. Cecile O' Rahilly: mono thaeth in fhirmimintni cona frossaib retland for dunignuis in talman no mani thi in fharrgi eithrech ochargorm for tulmoing in bethad no mani mae in talam...

5. Caesar, De Bello Gallico, i.l.

6. Diodorus Siculus, v.29-31.

7. Strabo, vii. 1.2.

8. Aristotle, fr. 610; Politics, vii.10.

9. Pliny, iii.57, quoting c.l.i.tarchus, who was there. Arrian, vii. 15.5-6, is inclined to discount it, 'given that no other people [than the Romans] was so possessed by hatred of despotism and its very name'.

10. Polybius, Histories, i.l.5.

11. ibid., vi.52.

12. ibid., vi.56.

13. Strabo, vi.1.2.

14. Pliny, Natural History, 29.1.7.-14.

15. Juvenal, vi.455.

16. Aulus Gellius, Noctes Atticae, xvii. 17.

17. Strabo, v.3.6.

18. Tacitus was right to cla.s.sify the Veneti and Fenni as neither Germans nor Sarmatians (who were Iranian nomads, related to the Scythians). But he goes on to identify the Peucini with the Bastarnae, known to have been Germanic (Strabo, vii.3.17).

19. Tacitus, Germania, xlvi.

20. Ptolemy, Geography, iii.5: 'katekhei de tebar;n Sarmatian ethne megista hoi te Ouenedai par' holon tn Ouenedikn kolpon'.

21. Strabo, vii.3.2, vii.5.2.

22. Lambert (1997: 123). These two were found in the regions of Nievre and Autun in France. The ordinal numbers from the potter's kiln in La Graufenesque are on p. 131.

23. Polybius, Histories, ii.17; Livy, v.34. Cf. Cunliffe (1997:71).

24. Martial, Epigrams, iv.60.8.

25. Lehmann (1987:76ff.).

26. Isidore, Etymologiae, xiv.6.6: 'Scotia idem et Hibernia proxima Britanniae insula, spatio terrarum angustior, sed situ fecundior. Haec ab Africa in Boream porrigitur. Cuius parles priores Hiberiam et Cantabric.u.m Oceanum intendunt, unde et Hibernia dicta ...'

27. Avienus, Ora Maritima, 11. 108-16: 'Ast hinc duobus in sacram, sic insulam / Dixere prisci, solibus cursi rati est. / Haec inter undas multa[m] caespitem iacet,/Eamque late gens Hiernorum colit./Propinqua rursus insula Albionum patet./Tartesiisque in terminos Oestrumnidum/negotiandi mos erat. Carthaginis/Etiam coloni[s] et vulgus inter Herculis/Agitans columnas haec ad[h]ibant aequora.'

28. ibid., II. 98-9: ' ...metallo divites/stanni atque plumbi ...'

29. Cunliffe (1997, ch. 8); Cunliffe (2001, esp. ch. 7).

30. They are detailed meticulously, and compared globally, in Gensler (1993).

31. Polybius, Histories, ii.17.

32. Reported in Cary (1954: 180).

33. Gildas, De Excidio Britonum, 6: '...ita ut in proverbium et derisum longe lateque efferretur quod Britanni nec in bello fortes nec in pace fideles'.

34. Tacitus, Dialogus de Oratoribus, x. 1-2.

35. Irenaeus, Adversus Haereses, i, preface.

36. Domitius Ulpia.n.u.s, Digest, x.x.xi.1.11.

37. Sidonius Apollinaris, Epistulae, iii.3.

38. Plutarch, Marius, fin.

39. Tacitus, Agricola, xxi.

40. Juvenal, Satires, xv.110-12.

41. Jackson (1994 [1953]: 107-10); Smith (1983).

42. Tomlin (1987).

43. Menendez Pidal (1968: 19).

44. Harris (1989: 315-16).

45. Augustine, De Doctrina Christiana, prologue 4.

46. Caesarius Arelatensis, Sermones, vi.l-2; viii.l.

47. Eutropius had written in the fourth century: 'Trajan, having conquered Dacia, had transferred there boundless numbers of people from all over the Roman world to tend the fields and the cities.' Breviarium ab urbe condita, viii.6.

48. Bourciez (1967: 30, 135-7).

49. The evidence is marshalled in Keys (1999, chs 13-16).

50. Weale et al. (2002).

51. Terrence Kaufman's calculation, using the standard Swadesh list of two hundred basic word meanings. Thomason and Kaufman (1988: 365).

8 The First Death of Latin

1. Monumenta Germaniae Historica, Scriptorum, i, 1.31.14.

2. This is quoted in Wright (1982:109), as at Vienna Nationalbibliothek 795. I have followed Migne (also quoted by Wright) in correcting sene to sine.

3. I am stating here as simple fact the thesis established with great doc.u.mentary effort by Roger Wright since 1982. The alternative would be to suppose that the p.r.o.nunciation of Latin had been kept constant for the preceding four centuries, without any special pleading or teaching. The experience in England since the Great Vowel Shift (fifteenth to sixteenth centuries) shows that scholars even of a written language that is quite distinct from their own do not, without copious urging and dispute, exert themselves to keep its sound system separate from that used in their daily speech.

4. De dissensionibus filiorum Ludovici pii, iii, ch. 5, dated by Studer and Waters (1924: 24) to 841-3. The text is there quoted in full.

5. Wright (1982: 124).

6. '...Et ut easdem omelias quisque aperte transferre studeat in rusticam Romanam linguam aut Thiotiscam, quo facilius cuncti possint intellegere quae dic.u.n.tur.' As quoted in ibid.: 120, 122, from Monumenta Germaniae Historica, Legum, iii, 2.1.

7. Menendez Pidal (1972: 24-5); also quoted in Wright (1982: 173).

8. Dante, De vulgari eloquentia, i.9.8-11: 'nec aliter mirum videatur quod dicimus, quam percipere iuvenem exoletum, quern exolescere non videmus: nam quae paulatim moventur, minime perpenduntur a n.o.bis, et quanto longiora tempora variatio rei ad perpendi requirit, tanto rem illam stabiliorem putamus. non etenim admiramur, si extimationes hominum, qui parum distant a brutis, putant eandem civitatem sub invariabili semper civica.s.se sermone, c.u.m sermonis variatio civitatis eiusdem non sine longissima temporum successione paulatim contingat, et hominum vita sit etiam, ipsa sua natura, brevissima. si ergo per eandem gentem sermo variatur, ut dictum est, successive per tempora, nec stare ullo modo potest, necesse est, ut disiunctim abmotimque morantibus varie varietur, ceu varie variantur mores et habitus, qui nec natura nec consortio confirmantur, sed humanis beneplacitis localique congruitate nasc.u.n.tur. hinc moti sunt inventores grammaticae facultatis: quae quidem grammatica nihil aliud est quam quaedam inalterabilis locutionis ident.i.tas diversihus temporibus atque locis.'

9. Dante, Convivio, i.2.9: 'Movemi limore d'infamia, e movemi desiderio di dottrina dare la quale altri veramente dare non pu.'

III Languages by Sea

9 The Second Death of Latin