On the Aesir Languages and the Boundaries Assigned to Them

The expression ‘the Aesir tongue’ is convenient upon a railway ticket and frequently disastrous in a learned work. It may signify the speech of a western herding clan, the language employed in an eastern provincial court, or an inscription whose author has been dead for twenty-five centuries. These are connected objects of study, but a connection does not confer identity. I have watched an interpreter, engaged upon the assurance that he ‘spoke Aesir’, negotiate the purchase of fodder admirably and then mistake a complaint concerning inheritance for an offer of marriage. The subsequent embarrassment was distributed unequally. A useful account must therefore distinguish historical descent from present comprehension, and both from the written conventions by which considerable differences of speech are concealed.

Our recorded history begins with Imperial Aesir, the language of the Old Northern Empire, whose surviving documents run from the eighth century before the era to the third century of the present one. Its evidence is scanty and inconveniently selected: we possess boundary inscriptions, dedicatory tablets, fragments of revenue accounts and passages copied by later scholars whose fidelity cannot always be tested. A language reconstructed chiefly from taxation will appear more orderly than its speakers deserve. We can establish much of its nominal inflection and several common verbal constructions; we cannot write a dependable conversation between two ordinary speakers. Professor Vigmar’s recently published imperial dialogue supplies precisely this deficiency, at the expense of supplying most of its evidence also.

Imperial nouns bear endings indicating their office in a sentence, and adjectives commonly repeat distinctions of number and grammatical class already expressed by the noun. The verb records person and number, although several endings remain disputed. Word order consequently permits arrangements which a Human reader may initially mistake for carelessness. In the better-preserved decrees, placing the recipient before the action appears to give that recipient particular prominence. I say ‘appears’ because decrees are poor witnesses to ordinary emphasis, and because one supposedly emphatic opening proved, upon my examination of the stone, to be the end of the preceding sentence.

The written record of Imperial Aesir tells us considerably less about pronunciation than its neat printed editions suggest. Words once spelt differently may have come to sound alike; a doubled vowel may indicate duration, a boundary between syllables, or a copyist’s error. Later readers pronounced ancient texts according to their own speech, much as a provincial lecturer contrives to make every historical period native to his birthplace. I have therefore omitted the customary account of the ‘majestic resonance’ of Imperial Aesir. No one has heard it. Its verse supplies evidence concerning syllables and perhaps stress, but majesty has yet to acquire a satisfactory method of measurement.

By the third century before the era the empire had divided into separate centres of administration, although the record permits us to distinguish Imperial Western Aesir from Imperial Eastern Aesir only from about the third century of the present era. These names cover the long period extending to about the middle of the thirteenth century. Political separation did not divide the language cleanly on the afternoon when two sovereigns first required separate treasuries. Early western and eastern documents share numerous forms, while provincial variation occurs within each dominion. The names describe traditions which became increasingly distinct through continued use, though they should never be taken to imply that every inhabitant obeyed a linguistic frontier with the punctuality expected of customs officers.

Imperial Western retained a conspicuous system of noun endings, grammatical classes and agreement. Its surviving narratives favour compact clauses, and its scribes delighted in compounds whose successive elements could specify an object with formidable economy. Written forms such as skeldr and vauth recur in early records with meanings connected respectively to an allotted portion and a binding pledge. Their exact vowels remain uncertain; the alternation of vowels within related words is better established: a change inside the stem could mark number or distinguish a verbal form, without obliging the writer to append another syllable.

Imperial Eastern nouns likewise declined, though their endings followed a different course from those of the west. Eastern texts make increasingly regular use of verbal prefixes, several of which distinguish an action considered in progress from one considered as reaching its result. The account books are particularly useful here. A steward had practical reasons to distinguish ‘was delivering’ from ‘delivered’, and an angry annotation beside an unpaid consignment has clarified a construction which two generations of grammarians approached through ceremonial poetry.

The last Imperial centuries, approximately the eleventh and twelfth, are poorly illuminated. This is a deficiency in our shelves; it gives us no warrant to imagine two centuries of conversational suspension. The later Dwarven administration preserved some local names and commercial formulas, but its clerks generally translated what interested them and ignored the grammatical particulars that interest me. Modern Tribal Western and Tribal Eastern become recognisable in the surviving record from about the middle of the thirteenth century onward. ‘Tribal’ is an inherited scholarly designation which I retain for convenience; both languages have long been spoken in settled communities, including towns whose inhabitants encounter a mammoth chiefly when it obstructs a crossing.

Present-day Tribal Western preserves enough of the older inflection to make a historical grammar immediately useful, though preservation must not be confused with immobility. Endings have merged in some districts and remained distinct in others; familiar compounds have become single words whose constituents speakers no longer recognise. In the western uplands, vauth survives in terms for obligations acknowledged before witnesses. At the coastal wool exchanges, its descendant may mean the paper recording such an obligation. A young clerk corrected my translation accordingly. I objected that the older sense was securely attested, whereupon she placed the document in my hand and asked whether I intended to file an obligation in the cabinet. I amended the entry.

Tribal Eastern has generally shortened or lost certain unstressed vowels preserved in its older writing, producing consonant groups which foreign primers make unnecessarily alarming. Several consonants have contrasting forms, one articulated with the tongue raised further towards the palate. These distinctions can separate words, and should be learnt before the visitor congratulates himself upon an intelligible accent. The inherited noun system remains substantial, while the distinction between continuing and completed action has become central to the ordinary verb. Eastern speakers are perfectly capable of describing time without specifying whether an undertaking succeeded; their grammar nevertheless obliges a learner to attend to the shape of an event more frequently than several Human languages do.

These descriptions are useful only if their geographical looseness is remembered. Tribal Western and Tribal Eastern each comprise overlapping varieties, and the frontier between them contains communities whose speech resists an economical label. Along the Kalvik winter route, neighbouring camps commonly understand one another with little difficulty, yet speakers brought together from its opposite ends may require repeated explanation. The road between them crosses no single point at which comprehension suddenly fails. Marriage, shared pasture and seasonal employment distribute forms unevenly. A settled market town may preserve an old local pronunciation while adopting the vocabulary of distant customers; a mobile clan may carry a remarkably stable speech through several provinces.

I attempted to measure these relations by asking speakers to repeat accounts delivered by visitors from other districts. My first results exaggerated mutual comprehension considerably: the listeners knew the expected account of a lost calf and supplied particulars they had never understood. When I substituted an unfamiliar dispute involving a damaged weighing machine, performance declined sharply. One herder accused me of making the test difficult on purpose. I admitted the charge and explained its necessity; he then proposed a second test in which I was required to identify individual calves from his descriptions. My failure helped distinguish unfamiliar subject matter from unfamiliar language, and he assisted with the revised procedure upon terms which included a generous account of my earlier performance.

Aesir longevity further complicates any map drawn from a single visit. Three living generations may preserve usages separated by more than two Human centuries, and an elderly speaker’s recollection can sometimes be checked against a document written in youth. Such testimony is valuable within its proper reach; it cannot bridge the interval to the Old Northern Empire, whose fall lies more than two thousand years behind us. I have opposed, in print and in a rather draughty clan hall, the assertion that an ancient song must preserve imperial pronunciation because its custodians are exceptionally long-lived. The custodian concerned understood my objection sooner than his visiting admirers did. He also showed me where his grandfather had altered a line to restore its metre.

Schools, railways and the telegraph are now spreading conventions which will make written intercourse easier without immediately making speech uniform. Western printers increasingly agree upon endings that several districts no longer pronounce alike. Eastern telegraph offices omit certain verbal prefixes and depend upon context, occasionally with expensive results. I supported the restoration of one such prefix after a message concerning an expected consignment was read as confirmation of its arrival; the railway’s objection concerned the additional chargeable syllable. An adequate account of these languages must consequently describe the written standards beside the speech of particular communities, giving place, speaker and circumstances wherever possible. In the chapters that follow I have retained variant forms even when they spoil a handsome table, and identified uncertain readings even where a confident choice would make the exposition easier. The reader wishing to learn a language for actual use should begin with the district to which he intends to travel, and should ascertain how much assistance his prospective teacher received from the pupil during their first conversation.