notes on the hermeneutics of early irish texts

Notes on the Hermeneutics of Irish texts

In reading translations of ‘old’ and ‘middle’ texts which are no longer intelligible to native speakers of the modern forms of the language it’s believed to be in, when you read of a character who needs “twelve heroes to lift up the hair about his eye with iron forks”? you should question the translation (imo it means someone had reaped a meadow and needed a dozen stout lads to toss the hay onto the wains with iron pitchforks hair = (ir) fear and appears to be a kenning. words for school and similar institutions are mistakable for words for eye where english and irish mingle.).

Currently accepted translations of texts in what scholars classify as Irish for example were made long after the language in which they were written had become unintelligible to existing Irish speakers, so they were sometimes done not by native speakers of old or middle Irish, but by scholars using the techniques of their times to reconstruct the extinct language, and not all of these were even native speakers of an extant Irish dialect. It was a very difficult undertaking, which could only ever be experimental. Much of it was done when there were so few scholars working in the field that they had few critics – their work was never tested or checked over at the time, and still hasn’t been, except tautologically (and therefore not validly), using the lexicography and grammar derived from the translation itself, not from any outside source.

Their methodology – the way they went about working things out – would not get them a pass at undergraduate levels in this day and age. Academic methodologies have advanced greatly since the promotion of scientific enquiry to academic levels in the 19th century.

These days, in any piece of research, a single experimental attempt and its results should have to show repeatability before it could be awarded the marks, or get funding. But the old texts of many languages that came to light during and after the Renaissance including the Irish texts, were translated long ago by methods that would no longer be considered sound in any other discipline. Where more than one translation was attempted, the winner prevailed in an academically insane one-only knock-out competition. Then chronologies have been got up for these texts on the basis of what would be now considered to be inadequate evidence. These have never been revised in the light of modern advances in hermeneutics. (Hermeneutics = the interpretation of texts.) Subsequent retranslations almost invariably use the lexicons and grammars derived from the established ones, which of course, tautologically ‘confirm’ not that the earlier translations are correct (such retranslations don’t test them at all), but only that their dubious lexicography and grammars are internally consistent.

A typical translation, usually done more than a century ago or based on work done then, proceeds more or less as follows:
• A text is selected for the attempt and a methodolgy devised.
• It is identified, dated, and contexts are noted, since data of this kind can shed light on the possible meanings of words.
• The language it is written in is identified as accurately as possible – this is based on observations of similarities its vocabulary and grammar might have with extant languages to which it may be related.
• An attempt to translate it is made.
• During this process, the words that most resemble words in extant relatives of the text’s language are considered to be close in meaning (usually), and these are entered into the lexicon. It should be noted that this process is not simple translation, but is already dependent on the comparative philology (as it was then called) of the time. Comparative philology assumed that if a word in one language resembles a word in another in both sound/form and meaning, they are probably related. As a corollary, the assumption can only be that if a word in one language resembles a word in a language that is thought to be related, it’s is possibly the same word and may have a similar meaning. As you can see there is considerable room for error here.
• Words that bear no clear resemblance to any words in any closely related languages are examined for resemblances to words in less closely related languages. Thus the translator must use the comparative philology of the time to find possible meanings and the scope for error takes a quantum leap! (The comparative philology of the 19th century was very naively done from a very unrealistically restricted database, by scholars who were still forging the rules, and even now, historical linguistics as it is now often called, has not responded to advances in linguistics made since De Saussure in the twentieth century. Much 21st century historical linguistics is still based on academically unsustainable assumptions about language change made before then. These educated guesses go into the lexicon. All this is nevertheless okay, as long as no one loses sight of the FACT that this is all hypothetical (i.e., controlled, educated guesswork.)
• Affixes and regular initial letter mutations etc (including grammatical inflections) are sought and examined and from observations of them a grammar is deduced. If the lexical guesses are accurately recorded, and this deduced grammar is accurately described, any attempt to translate the text using that lexicon and that grammar should produce (or be able to produce) the same translation every time. If it does, then the work is internally consistent and therefore a good attempt – provided that the translation makes sense. So such an experiment is good in itself, but it can only provide us with one scholarly opinion. Before it can be claimed as knowledge, it must be validated – it must be checked for repeatability by other scholars. Later, more modern experimenters must see whether, using modern methods, they obtain the same lexicon and the same grammar. To use the lexicon and grammar derived from an earlier experimental translation to check that translation is tautological – it will always prove it correct, but it will not tell us whether the lexicon itself is correct, nor the grammar.

This independent checking has NOT been done in the case of most of the Irish texts, as far as I know. If it had been, the Dictionary of the Irish Language and Thurneysen’s Grammar would no longer be in use.

For best results, double-blind conditions should apply. That means that those who make other attempts should not consult the work of those who have made previous attempts.They must start with the same equipment as the other experimenters: the text and all the academic resources of their own time, and a mind as untainted as possible by the existing dogmas. Those who gave us the translations we have now compiled their own lexicons and derived their own grammar, and those who check their work must likewise build their own without referring to that of their predecessors in the field.

Perhaps a 21st Century test for repeatablity would look like this.
• The text is selected given to six teams of experts in the field. (In double blind experiments these teams do not consult each other’s work at all, or the work of previous workers in the field.) These experts must know well all extant dialects of Irish (native speakers should be chosen, and must also have a good knowledge of all languages that may be related to the text – I would consider other modern Celtic languages, English, Dutch, Danish, Icelandic, German, French, the Spanishes, Moorish and Portuguese to be essentials. They don’t have to speak them all, but should have closely examined their vocabularies and grammars. They should be well acquainted with other old texts, bearing in mind that their previous translations, lexicons and grammars are the result of unrepeated, therefore unverified guesswork, and so cannot be used to support their own guesses as that would be a violation of the double-blind. They may however use their own interpretations of them, done under similar double-blind conditions. These would include the old English mss, the Gothic, Flemish, Icelandic, etc texts and even Sanskrit and Hebrew would be relevant. They must also have had sound training in postmodern hermeneutics, which keeps a sharp look-out for personal biases, political assumptions, subtle ‘spin’ and other distortive psychological factors. The translations we have are imo badly distorted by political, cultural and ecclesiastical assumptions held by the translators and the dominant culture of their day.
• A 21st century methodology is devised – each team builds its own. Advances in methodology since the existing translations were made have been very considerable. Hermeneutical training these days teaches the translator to identify and eliminate their own personal, political, ecclesiastical and cultural biases (or at least declare them). Not revising the old translations in the light of those advances is like insisting that the world is flat on the basis of 12th century church dogmas.
• Its identity is checked carefully, an attempt is made to decide on the age of the MS, and all relevant contexts are noted, since data of this kind can shed light on the possible meanings of words. Strict veracity must be maintained. This has not yet been done. The science of dating mss is a bit of a vicar’s egg – very good in parts, but most of us would forgo the egg. In 21st century work, sometimes the correct answer is ‘we don’t yet know’. A translation with a lot of gaps in it is sometimes a more truthful and accurate one than one without – especially when the meanings are odd or meaningless. Similarly, where a team of experts work together on a single text, whenever their opinions differ, the final translation should show all the opinions of the group, along with their reasons for them.
• The language it is written in is identified as accurately as possible – this is based on observations of similarities its vocabulary and grammar might have with extant languages to which it may be related.
• An attempt to translate it is made: where there is disagreement within a team, these should be declared in the final translation.
• During this process, the words that most resemble words in extant relatives of the text’s language are considered to be close in meaning (usually), and these are entered into the lexicon. It should be borne in mind that modern Irish speech has been contaminated by the introduction of words from the lexicon, which has been derived unscientifically. These words are usually easily enough recognised by sensitive scholars. Errors are possible here, but usually not too bad.
• Words that bear no clear resemblance to any words in any closely related languages (dialects of modern Irish) are examined for resemblances to words in less closely related languages using the comparative philology of the time. (other celtic, then english, icelandic etc . Errors are more likely here, and can be minimised by leaving gaps (when the translator has no idea) or offering several opinions (when the translator considers several possibilities, or when two team members disagree). These educated guesses go into the lexicon.
• Affixes and regular initial letter mutations etc (including grammatical inflections) are sought and examined and from observations of them in the light of a deep and detailed knowledge of the way such indicators of grammar are used in other languages, a hypothetical grammar is deduced.

Now we have not one scholarly attempt at a translation, but seven, one done in the past when the hermeneutical sciences were in their infancy, the other six using state of the art hermeneutics.

Now comes the test. Compare the results. Where all seven agree as to the meaning of a sentence, it might be safe to say that they are probably correct – but not that they are correct. When most of them agree but one disagrees, the certainty is less – more research needed – and gets less the more they disagree. Perhaps they all agree fairly well on the words that occur also in Irish, but you could expect an array of opinions, even within a group, on words not recognisable as Irish, and these should all be declared.

To put it simply, if all six teams come up with a lexicon that exactly resembles that which is included in the DIL, plus an exact replica of Thurneyson’s grammar, they might have a case for maintaining that the existing translations are accurate. Otherwise, they have not. And they have not until such a test has been done. All such conscientious, academically sound attempts must be given equal credence the old ones, done according to out-moded methodologies, using 19th century comp.phil (or older) should be regarded with acute skepticism.

It doesn’t stop there. Once you have a good variety of opinion, swarms of scholars working over them all  comparing them minutely, observing where they all agree (probably right) and where they split into two, three or more camps (any or none could be right) would make dramatic progress towards a much more realistic translation: decades of work could be done. Where difference of opinion occurred, more research might resolve the problem. New translations working from their own lexicons would continue to arise, and my guess is that opinion would soon begin to converge on much more realistic interpretations of the old texts than we now have. But while a highly trained Oxford scholar ( I name no names) can say blithely, ‘There is only one opinion in Celtic Studies’ we may as well all pack up and go home!
(In an article this short there are bound to be some over-simplifications, but I believe I am making an essentially valid statement.)

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s