Is It English or Engelsk?—part 2
Let’s now consider the more fundamental change from the Object-Verb (OV) to the Verb-Object (VO) pattern that characterized the transition from Old English to Middle English. It too cannot be attributed wholly to the Vikings for several reasons. First, the VO pattern is attested in Old English alongside the more frequent OV pattern (the Old English examples are taken from The Homilies of the Anglo-Saxon Church, containing Sermones Catholici or Homilies of Ælfric, composed in the late 10th or early 11th century):
OV in Old English:
|‘With two things God had endowed man’s soul.’ [ÆCHom]|
VO in Old English:
|‘That he would not have taken mankind forcibly from the devil.’ [ÆCHom]|
Significantly, this example of VO order in Old English contains the verb niman ‘to take’, which in due course was driven out of the lexicon by a Norse loan taka of the same meaning (the old root niman was preserved in nimble originally meaning ‘quick to grasp’ and numb originally meaning ‘taken, seized’).
Second, the shift to VO in English cannot be characterized as borrowing a grammatical pattern from Old Norse because the latter was itself a mixed-word-order language, with both OV and VO as possible orders (curiously, Faarlund in his book Syntactic Change: Toward a Theory of Historical Syntax, p. 110, calls Old Norse a “free word order language”, highlighting the non-uniformity of its word order). For example, Rögnvaldsson (1996, p. 56) provides the following examples of OV and VO order in Old Norse:
OV in Old Norse:
|‘But I believe I have not seen this man.’|
VO in Old Norse:
|‘Lyting from Samsstadir will have killed him and his brothers.’|
In fact, Old Norse, much like Old/Middle English, was transitioning from the OV order, which was found in the Proto-Germanic, to the VO order, which characterizes all Modern Scandinavian languages. While it is difficult to estimate the relative frequency of different constructions in an earlier form of a language, it appears that in the relevant period the language of the Scandinavian settlers in Britain was more predominantly OV rather than VO (this issue is discussed in great detail in Jens Haugan’s doctoral dissertation Old Norse Word Order and Information structure).
Thus, the change from OV to VO derives not from the introduction of a new, previously unavailable structure, but rather from the changing frequency of the two existing patterns, which happened gradually during the Middle English period (11th-15th centuries). As the following example from Chaucer’s Malibee shows, the OV pattern was still in use in 1330s:
OV in Middle English:
|‘I can keep and defend myself and my house in such a way.’|
A closer look at the changes in the order of the major sentence parts—subject, verb, and object—reveals that Faarlund and Emonds’ claim that “wherever English differs syntactically from the other Western Germanic languages […] it has the same structure as the Scandinavian languages” is factually incorrect. One obvious difference between English and, say, German is that in the latter verbs must appear in the second position in the sentence (linguists refer to this pattern as V2 for ‘verb second’), regardless of whether the first position is occupied by the subject or by some other phrase, whereas in (Modern) English this rule does not apply, as can be seen from the English translation of the following German sentence:*
V2 in German:
|‘Yesterday, Martin has read the book.’|
Note that the English pattern (Gestern Martin hat …) is ungrammatical in German. The problem for Faarlund and Emonds is that V2 is found not only in West Germanic languages (German, Dutch, Yiddish, Afrikaans, Frisian), but in all North Germanic languages as well, including Norwegian (illustrated below) and Icelandic:
V2 in Norwegian:
|‘Yesterday, Martin has read that book.’|
The only instances of V2 in English are found in sentences that start with a negative adverb such as never or phrases containing only, where the verb must appear in the second position. (This is particularly apparent if the verb in question is an auxiliary or a modal rather than a lexical verb.) Consider, for example, [Never] have I seen such a creature! and [Only here] have I seen this! Unlike Modern English, Old English was a well-behaved Germanic language with the V2 pattern, illustrated by the following sentence (repeated from above):
V2 in Old English:
|‘With two things God had endowed man’s soul.’ [ÆCHom]|
Obviously, the change from V2 in Old English to no-V2 in Modern English cannot be explained by copying a Scandinavian pattern. On the contrary, the V2 pattern persisted longer, well into the 13th century, in the northern areas where Scandinavian influences were most pronounced, according to Fischer and van der Wurff’s chapter on syntax in A History of the English Language (2006, edited by Richard Hogg and David Denison, p. 184). Thus, they find explanation for the loss of V2 in English not in contact with the Scandinavian-speaking Vikings but in complex structural properties found in Old English itself. Yet the Vikings in the north (and later also the Normans in the south) had a role to play in the gradual disappearance of the V2: they simply got rid of patterns that were too complicated for adult second-language learners of English to handle. In particular, Old English had different rules for placement of subjects depending on whether they were nouns or pronouns: if a subject was a pronoun, it was placed before the verb, and the V2 rule was not applicable. Old Norse, however, bestowed no special status on pronominal subjects, so Norse-speaking Vikings, who had no use for extra grammatical complexity in a language that was not their native one, gravitated towards the “one rule fits all” system, extending the preverbal placement of subjects from pronouns to nouns.
A similar reasoning can be applied to explaining other grammatical changes, such as the decline of the case system and the verbal conjugation, during the transition from Old to Middle English: Norse-speaking Vikings and later French-speaking Normans eliminated those aspects of the Anglo-Saxon inflection that they found too difficult. For example, the system of case markers—typically, indistinctly pronounced endings on nouns and adjectives—was too tricky for non-native adult learners to pick up. As a result, they simplified the system, first by using certain endings in the wrong places and then by dropping the case endings altogether. Consider, for instance, the inscription on the Kirkdale sundial, embedded in the wall of Saint Gregory’s church in North Yorkshire.
The inscription is written in a late form of Old English, as it dates from 1055-1065, judging by its reference to “the days of King Edward and Earl Tostig” (in the bottom two lines on the right panel): King Edward is Edward the Confessor, who reigned in 1042–1066, and Earl Tostig is Tostig Godwinson, the son of Earl Godwin of Wessex and the brother of Harold Godwinson (who opposed William the Conqueror in the Battle of Hastings); Tostig held the Earldom of Northumbria in 1055-1065.
That Scandinavian speakers had something to do with this inscription is obvious not only from its location in the heart of Danelaw (note that kirk- in the toponym is a Norse word for ‘church’), but also from the text itself. The two personal names that appear in the bottom line of the central panel, “Hawarth” and “Brand”, as well as the two names in the first line on the left panel, “Orm” and “Gamal”, are Scandinavian in origin, reminding us that language contact often results in expansion of the repertoire of names as well as of the lexicon.
Also of Scandinavian origin is the use of the patronymic formula Gamal svna, common in Scandinavia and rare in Old English outside of Scandinavian-influenced regions. This formula gave rise to English last names such as Johnson, Stevenson, and Peterson (the equivalent English patronymics end in -ing, as in Browning). The use of the word eorl ‘earl’ as a title is another sign of Nordic influence: in Old English the word tended to be used in poetry with a general meaning of ‘man, warrior, hero’, while the rank of ‘earl’ was expressed by another Old English word, ealdormann (which survives in Modern English as alderman). The word solmerca for ‘sundial’ is half Old Norse and half Old English: the first element of the compound, sol- ‘sun’, is Norse, whereas the second element is likely native to English. Finally, the word tide ‘time’ is likewise of Scandinavian origin; in Modern English it is preserved in its earlier meaning of ‘point in time’ only in expressions such as Yuletide.
The inscription on the Kirkdale sundial also exhibits examples of the loss of inflection in late Old English. Consider, for instance, the highlighted image of the sundial’s central panel; here, the dative adjectival ending ‑um, originally used only for masculine and neuter adjectives, is extended to a feminine adjective agreeing with tide ‘time’, replacing the ending ‑e used in such forms in earlier stages of the language. It has been suggested that contact between speakers of English and speakers of Norse was crucial for this process leveling inflectional morphology. According to Matthew Townend,
“English and Norse […] were probably mutually intelligible languages […] However, while cognate English and Norse words were generally similar, or even identical, in their basic form the one aspect in which they often differed was their inflexional [sic] endings: compare, for instance, Old English giest and Old Norse gestr (‘guest’), or guma and gumi ‘man’, or scipu and skip (‘ships’). In a situation in which speakers of the two languages were repeatedly in contact with one another, on a daily or even a domestic basis, it is quite possible that these inflexional differences became eroded or ignored, as they played no role (or were even a hindrance) in effective communication between speakers of the two languages. […] hence [inflections] decayed…” (The Oxford History of English, pp. 82-83)
A similar point is expressed by Albert C. Baugh and Thomas Cable:
“in many words the English and Scandinavian languages differed chiefly in their inflectional elements. The body of the word was so nearly the same in the two languages that only the endings would put obstacles in the way of mutual understanding. In the mixed population that existed in the Danelaw these endings must have led to much confusion, tending gradually to become obscured and finally lost.” (A History of the English Language, Fifth Edition, 5th edition, p. 104)
The Kirkdale sundial inscription also illustrates the development of the English genitive. Like other cases, in earlier forms of Old English genitive was marked by special endings, but in the later period exemplified by this inscription, the genitive endings disappear. Hence, we see Eadward dagum cyning (the latter is abbreviated as CNG) and Tosti(g) dagum eorl, instead of the expected earlier formsEadwardes dagum cyninges and Tostiges dagum eorles, respectively.** Which noun denotes a possessor and which one the possessed has to be understood purely from the word order and the context (the postposition of the titles ‘king’ and ‘earl’ is typical of medieval writing). So where does the modern possessive marking ‘s come from? Instead of being a Scandinavian import, as Faarlund and Emonds claim, it is a remnant of the genitive case ending that survived in southern parts of England less affected by Old Norse. It is a reduced form of the earlier ending -es originally used for certain masculine and neuter nouns, but later extended to feminine nouns as well (the genitive ending for plural nouns was -as, which too lost its vowel over time).
Nor is the phrasal genitive, that is the attachment of ‘s to the phrase rather than the head, a Scandinavian structure. First, the appearance of phrasal genitive in English dates to a much later period, centuries after the Viking Age. Even in texts like the ballad “Sir Patrick Spens”, which relates historical events of the late 13th century, we still find the ‘s attached to the head rather than the whole phrase, as in the King’s dochter of Noroway for ‘the King of Norway’s daughter’. Second, Old Norse did not have phrasal genitive, so it could not have contributed it to what became Modern English (e.g. Icelandic does not have phrasal genitive to this day). The phrasal genitive developed independently in English and in Norwegian and made its appearance in both languages relatively late. At least in English, the phrasal genitive is attributable to the so-called “his genitive”, a construction particularly popular in the late sixteenth and early seventeenth century and employed in the titles of John Lyly’s Euphues His England (1580), Ben Jonson’s Sejanus His Fall (1603), and John Donne’s Ignatius His Conclave (1611). Over time, the possessive pronoun his wore down to a mere ‘s, making it indistinguishable from the remnants of the genitive case ending, and its use was extended to feminine nouns. The very same development happened in another West Germanic language, Afrikaans, where the possessive pronoun sy ‘his’ developed into an all-around possessive element se, which can appear with both masculine and feminine possessors, as in die man se kinders ‘the man’s children’ and die vrouens se kinders ‘the women’s children’.
As can be seen from the above discussion, Faarlund’s claim that these types of grammatical structures “are very unlikely to change within a language” is incorrect. In the next post, we will consider the more general question of whether grammatical patterns such as these can be borrowed from one language into another.
*In most Germanic languages—with the notable exception of Icelandic and Yiddish—the V2 phenomenon happens only in main clauses. For clarity’s sake, we will ignore what happens in subordinate clauses.
** The form eorles is attested, for example, in an inscription on an early eleventh century grave-marker from the Old Minster, Winchester, discussed in The Oxford History of English, p. 76.
Rögnvaldsson, Eiríkur (1996) ‘Word Order Variation in the VP in Old Icelandic’. In: Working Papers in Scandinavian Syntax 58, pp. 55-86.