Catlike Canines

Keep in mind this isn’t always the case but it seems in the case with at least a few communities in the entire world (some Amazonian/South American communities as well as the Akan in general) the word for dog’s derived from either cat or some other felid. It’s like how some Amazons refer to dogs as house jaguars or among some Akans, the word for dog is okraman and the word for cat’s okra (also soul).

It’s not that dogs haven’t been around in Africa and South America for a long time but possibly shorter than what’s expected in Eurasia, which’s where they’re first domesticated there. If wolves aren’t native to Africa at all, then dogs might inevitably be introduced and count even be a proper invasive species there (dog predation on livestock and monkeys have been noted before).

They’re even considered as such in some Latin American circles. Again not always the case but if/when dogs are recently introduced in South America and to a lesser extent, Africa then sometimes the word for dog may be likened to cats as a point of reference/familiarity. The fact that cats and dogs are sometimes closely entwined in witchcraft beliefs (which makes more sense with goats being victims, that dogs do prey on them) makes it a good semantic and semiotic association.

It’s like this study on one Ivorian community where witches have familiars taking on guises of dogs, leopards and cats and prey on goats. Likewise in another study, though this might not be true for all Amazons, witches are said to appear as dogs, jaguars and aeroplanes. This is what I’m talking about. Whilst not always the case either, if witches are likened to predators like dogs and leopards then this shouldn’t be surprising.

The fact that dogs can take on solitary behaviours, cats with some degree of social behaviour and some households have both cats and dogs made to hunt pests should make it a very unsurprising association.

The closest to the old school

Scots is like English in the same way Portuguese’s like Spanish. Very similar but also different. I actually regard Scots as the more Germanic counterpart to English. It’s not entirely free of Romance influence but it did preserve some features lost in contemporary English just as some Scots dialects (most notably Doric, Orcadian and Shetlandic) are heavily influenced by Nordic languages. This makes sense as Scotland itself was subjected to Nordic influence before, moreso with Shetland and Orkney as they used to be Norwegian territory!

It can be considered a form of colonial lag in that both English and Scots derive from Old English but only Scots has preserved the more archaic or old-fashioned features and moreso with some dialects. The fact that the Normans never invaded Scotland, so whatever Latin/Romantic influence there is in Scotland is somewhat more modest than with English at the very least. Scots isn’t necessarily entirely free of Romance influence but closer to its Germanic roots than English is.

The separation into Irish and Scottish Gaelic

Admittedly I’m not a linguist but whatever possible dialectal variations within Old Irish’s been speculated by some scholars. Whilst not always the case, it does explain some things. It’s like if Southern Irish dialects (especially in Kerry) treat gorm, the Irish word for blue, to encompass green vegetation but other Irish dialects don’t (as far as I know and excerpted, the Belfast glossary as well as the Rathlin dictionary and the 1768 dictionary, have gorm normally mean blue and sometimes black). If true, then it might be analogous to what became of Vulgar Latin.

There could’ve been dialectal variation within Latin, moreso when it comes to colonised people speaking their mother languages from time to time. It’s even speculated that Gaulish may’ve influenced both French and Breton. Logically, Spanish got influenced by Arabic. France’s got a substantial Celtic (though Brythonic) speaking population in Brittany and Spain used to be influenced by the Arabs. This might not be unique to both of them either. Scottish Gaelic could’ve been influenced by any pre-existing Celtic language in Scotland.

Similar things can be said of Irish to some extent. Whatever the circumstance, dialectal variations within Old Irish and Vulgar Latin’s enough to beget contemporary Italian, Irish, Scottish Gaelic, French, Spanish and Portuguese.

Historical Dialectology (Excerpt)

Remarks on the question of dialects
in Old Irish
Classical Old Irish is normally presented as a language without
significant dialect variations. In a well-known and influential paper,
Greene (1977:11) has put the matter thus: “Since the beginning of the
scientific study of the Celtic languages in 1853, when Zeuss published
his Grammatica Celtica, the term Old Irish has been applied to the
remarkably consistent form of the language found in Irish glosses
written into Latin manuscripts during the eighth and ninth centuries.”
Before going any further, something needs mentioning about the
nature of these glosses. The main bulk of material now known to have
been written during the Old Irish period of ca. 600 — 900 AD
(Thurneysen 1946: 1, 673 n. 1) is found in three main collections of
glosses, which are:
I. The glosses (Wb.) on the Latin text of the Pauline Epistles, as found
in a manuscript preserved in Würzburg. Three different hands appear
to have been at work, at different times, but the main body1
of glosses
has been ascribed (Thurneysen 1946: 4) to ca. 750 AD.
II. The glosses (Ml.) on a Latin commentary on the Psalms, as found
in a manuscript preserved in Milan. It seems possible2
to date them
around the earlier half of the eighth century.
III. The glosses (Sg.) on Priscian’s Latin grammar. The manuscript
itself has been dated to around 845,3
but there are indications that
certain portions of the material had been copied from much earlier
sources; this naturally complicates dating these glosses from a
linguistic point of view.
Furthermore, there are numerous other smaller collections, most4
which have been edited by Stokes and Strachan in their Thesaurus
Palaeohibernicus. The aim of this work was, as the editors (1901: xi)
themselves stated: “to facilitate the study of the interesting and
difficult language commonly called Old-Irish, and for this purpose to
24 Anders Ahlqvist
put scholars in possession of trustworthy materials in a convenient…
combination.” They also pointed out that: “in forming a collection of
texts on which scholars may rely with confidence the only safe rule is to
exclude all matter not found in MSS. anterior to the eleventh century.”
Some years later, Thurneysen published his first grammar (1909) of
Old Irish, which was in a very large measure based on the materials in
the Thesaurus. Almost forty years later, an English edition (1946), with
many additions and corrections, was published. This still serves as the
indispensable standard reference grammar for all serious students of
Old Irish.
At the same time, much material from the Old Irish period has been
recovered from manuscripts much later than those that Stokes and
Strachan would have judged acceptable for the purposes of the
Thesaurus. In editing these, the problems relating to ancient forms of
language preserved in much later manuscripts have become very
evident. Until quite recently, the main principle adhered to by most but
not all editors of such materials has been to try to reconstruct the state
of language as it was at the time when the text was assumed to have
been written for the first time. For this, the language of the Old Irish
glosses has very naturally served as a model, allowing for the retention,
in modern editions, of deviant forms, when and usually only when
editors have seen them as more archaic (i.e., chronologically older)
than their correspondents in the language of the Old Irish glosses.
Many of these materials are extremely important, and it is no wonder
that when we may (Greene 1977: 12) “speak of Classical Old Irish we
usually include reconstructions of this kind in the corpus, which thus
becomes imposing in size and extent.”
At this point, we may turn to the subject-matter proper of this
paper: what about the question of Old Irish dialects? Naturally, it
seems unreasonable to assume that there were no dialect divisions of
any kind in Old Irish, given that the language was spoken over a fairly
wide area, extending, as would seem rather likely, all the way from
Caithness in the North to Kerry in the South, covering much of
Scotland and (or so we must assume) all of Ireland. In a recent
interesting article (P. Kelly 1982), an attempt has been made to ascribe
the origin of certain highly-marked items of Old Irish vocabulary to a
certain dialect. Although the present paper will make no attempt to
deal with lexical matters, her general comments (p. 86) are well worth
repeating here: “Im Altirischen dagegen sehen wir uns mit einem
verblüffend einheitlichen Sprachsystem konfrontiert, verblüffend im
Dialect parochialism: the Scottish vowel length rule 25
Vergleich zu den frühen Stadien der Textüberlieferung anderer
europäischer Sprachen; ein Sprachsystem also, wo weder in der Lautund Formenlehre noch in der Syntax regionale Unterschiede bisher
festgestellt worden sind”. Also, and in a most helpful fashion, she has
looked for references to the matter in Thurneysen’s (1946) grammar,
with these results: “In seiner ganzen Altirischen Grammatik zieht
Thurneysen nur fünfmal, und dann nur zögernd, die Möglichkeit von
dialektaler Variation in Erwägung (Thum. Gramm. § 16, § 166(b),
§ 485)”. Naturally enough, Thurneysen’s comments will provide a
suitable point of entry to our little investigation, and his first three
references (1946: 12) to dialects are worth quoting in extenso:
Linguistic differences in the Old Irish sources are almost all
differences of period, and are the result of morphological development. Contemporary divergences, such as would point to dialectal
peculiarities, are very rare; cf. for instance the superlative in -imem
(§ 371) found only in the Milan glosses, or the varying forms of the
preposition air- er- ir- aur- (§ 823), between which, however, no strict
line of demarcation can be drawn; further the almost complete
absence of on, by-form of son ‘that’ (neut. § 479), in Sg. The paucity
of the sources does not suffice to explain this comparative uniformity; in the literary language a levelling and intermixing of dialects
must have taken place. This process was undoubtedly assisted from
the earliest times by the wandering poets, singers and scholars, who
would naturally wish to be understood everywhere. Further, in the
monastic communities of the sixth and following centuries, from
which our sources are ultimately derived, the teachers were drawn
from various parts of the country.
As will be seen below, his general comments are apposite enough,
but the specific instances he gives are not very useful for our purposes.
Regarding the “double” superlative ending in Ml., this might have
been quite useful for ascribing these glosses to a particular dialect area,
had it not been for the fact that the pattern of making superlative forms
(with endings distinct from those of the comparative) disappears quite
early, leaving no traces in the later language. In the case of air- etc.
‘before, for’, we seem, at first, to be on somewhat firmer ground, even if
it turns out that no firm rules can be established. As Thurneysen (1946 :
497) remarks: there “are two main forms of this preposition in our
sources: air with palatal, and er or (in Wb.) ir with neutral r\… They
often interchange in the same word and sometimes in the same text.
26 Anders Ahlqvist
Examples: air-dire ‘conspicuous’ Sg., irdirc Wb., erdairc Ml.” Somewhat more helpfully, he adds that this “variation is limited only by a
decided preference for er in Ml., and for air in Sg.” On the other hand,
if we may judge by the modern spelling of the word given by
Thurneysen, Irish oirirc and Scottish Gaelic òirdheirc, the same
labialisation seems5
to have taken place all over the Gaelic area, so that
I cannot draw any conclusion from this for the moment, at any rate.
The matter of the spelling of the anaphoric pronoun on, son and its
distribution (4 χ on / 80 χ son), on the other hand, seems quite relevant
in this context: as Hessen (1914: 1—2) has established, there is a
significant difference between Ml. and Wb. on the one hand and Sg.,
on the other, arising out of “die in Ml ganz, in Wb fast ausnahmslose
Anwendung der s-losen Form in der Stellung nach einer andern mit
diesem Konsonant auslautenden partícula augens”. Thus, he concludes as follows: “Während SG bei 80mal belegtem són nur 4 ön
aufweist, steht in Ml die letztere Form zur unlenierten bei der hohen
Zahl von zusammen nahezu 700 Belegen im Verhältnis von 3 zu 4, eine
Erscheinung, die ich auf dialektische Unterschiede in der Sprache der
beiden Codices zurückführen möchte”. Unfortunately, however, these
dialect differences seem to have left no trace in the modern language, so
that their geographical implications remain unknown to me.
Thurneysen’s fourth case (1946: 104) of possible “dialectal differences” has to do with the rise of palatalisation in Irish, allowing for
doublets like Wb. 9d
5 cuicce and Sg. 14b
4 cucae ‘to her’. In this case, the
Sg. form is representative of an older state of affairs, according to
Greene’s (1974: 135) account of these developments. The fifth case
(Thurneysen 1946: 306), finally, has to do with the variation between
initial/- and c- in words meaning ‘self’. It is now known (O’Rahilly
1976: 255; Byrne 1982: 167-169) that this variation was originally
one between forms of different meanings, something that obviously
complicates the issue, even if it does not exclude the possibility of
dialectal variation.
To summarise the discussion of what Thurneysen (1946) has had to
say about possible dialectal variation in Old Irish, the conclusion has
to be that he used this as a sort of pis-aller, to describe phenomena for
which he had no other explanations. However, it seems to me quite
reasonable to assume that future research may yet validate at least
some of his suspicions. In the case of the rise of palatalisation, for
instance, the fact of the historically older form cucae being found in
Sg., when the (as normally assumed) older Wb. glosses have younger
Dialect parochialism: the Scottish vowel length rule 27
cuicce may be explained in one of two ways: either the Sg. form is there
because it had been copied from an exemplar written in language more
old-fashioned than that of Wb. or it derives from a more conservative
dialect. Similarly, in an interesting discussion (Mc Manus 1983:
70 — 71) of how Lat. spiritus was borrowed into Irish, allusion is made
to “the fact that Ml. and Tur. preserve … the old nom. and
(spiurt and spiurto), while the Wb. glosses, which are older, use the
more modern forms (spirut, spiruto, spirito but also spirto),” which
“may simply be due to dialectal variation or a regional delay in
spirut/spiruto becoming the accepted standard.”
In Modern Irish and Scottish Gaelic, there is a rather important
dialect marker, which is that provided by the negative particles ni and
cha (cf. O’Rahilly,2
1972: 293; Ó Buachalla 1977, 1982: 431, and Ó
Dochartaigh 1976): the former is current in Ireland and the latter in
Scotland, while both occur side by side in Ulster Irish. The corresponding Old Irish forms are ni and nicon, which occur (DIL N:42.37 — 50;
46.15 — 21) in the three main collections of glosses, so that only a
detailed examination of their relative frequencies would determine
whether they really correspond to dialect differences. However, I shall
not be surprised if Wagner (1986: 1) yet turns out to be right in
assuming that nicon never was common in Southern dialects, even if it
was regular enough in the Northern-based standard language.
With this, we come to a crucial concept in any description of Old
Irish, namely that of a standard language. In a recent article (McCone
1985:102), it is cogently argued that our corpus of Old Irish is, to some
extent, a mixture of registers :
in extant contemporary Old Irish texts, to all intents and purposes
the Glosses, the overwhelming majority of usages conforms to a
conservative literary standard and, at least arguably, associated
register of speech, while sporadic deviations from this are mostly due
to occasional lapses into a basically sub-literary register approximating to popular speech.
Now, one fairly typical feature of some conservative registers of
written language is that they are felt to be so to speak supra-dialectal,
even if what that really means is that a particular dialect has, usually
for non-linguistic reasons, acquired a special prestige lacking in others.
Features identified as dialect ones, on the other hand, are more keenly
felt in the popular register. This brings us to a not unlikely dialect
28 Anders Ahlqvisî
feature that McCone (1985: 96 —97) has observed in the syntax of the
It concerns the construction in which a preposition governs a
relative clause. In Modern Irish, the normal construction is roughly the
same as in an English sentence like the man who(m ) I gave the book to,6
whereas Old Irish prefers the equivalent of the man to whom I gave the
book. However, there are a few isolated cases of the former type even in
Old Irish, namely (Thurneysen 1946: 322) Ml. 87d
l 5 nech suidigther loc
daingen dò ‘anyone that a strong place is assigned to’ and Sg. 26b
nrfail ni nadtai mo dligeth-sa fair ‘there is nothing that my law does
not touch on’.7
These McCone describes (1985: 97) as a “type of
broadly southern origins” that “was apparently confined to colloquial
usage for centuries and only cropped up occasionally in the literature”.
Accordingly, he argues fairly convincingly that “a northern locale for
development of a written Old Irish standard would be indicated”.
Therefore, “east Ulster, perhaps one of its great monasteries such as
Bangor, would have considerable attractions.”8
In this context, it is
relevant to note that, although basing her findings on totally different
evidence, Kelly (1982:89) comes to the following conclusion: “Daß der
politische Schwerpunkt nach 700 im Norden lag, hat wohl seinen
Niederschlag in der Standard-Schriftsprache gefunden, die somit im
wesentlichen nördlich orientiert sein dürfte.”
To corroborate the findings of these two scholars, I may add two
other facts relating to relative constructions in Irish and Scottish
Gaelic. In Old Irish, it is clear that the initial mutations found in special
relative forms of simple verbs are a secondary phenomenon. I shall
leave the historical reasons for this aside here,9
merely restating
Thurneysen’s description (1946: 315, 319) of the facts: lenition is not
found in Wb., only in special cases in Ml. and is widespread in Sg.;
nasalisation is usual but not universal in Wb., but more regular in
“later sources”, i.e., Ml. and Sg. This naturally leaves one with the
impression that initial mutations had become the rule in these cases in
later Irish. However, there are exceptions, as I have pointed out (1978 :
77 n. 10 and 1987a: 11 — 12) elsewhere: these turn up in present-day
Donegal Irish, as in is mairg a bios gan car aid ‘unhappy is the one who’s
without a friend’ and in an early fourteenth century grammatical text
from Co. Tyrone: cá méid críochnaigheas i «-e? ‘How many [names of
consonants] end in eT. To me, at any rate, it seems likely that we are
dealing with some sort of either historical or dialectal continuity in the
case of the pattern found in the glosses. If the former is the case, the
Dialect parochialism: the Scottish vowel length rule 29
non-mutated northern forms just mentioned are retentions of a pattern
that was on the way out in standard Old (etc.) Irish even as early as the
Wb. glosses.
The second feature concerns the use of adverbials in cleft sentences.
In Old Irish, they are normally followed by a non-relative verb (a
pattern that I have described elsewhere (1978: 70 — 71)). In the later
language, including a few isolated forms in Ml, and Sg., a relative form
of the verb is found, as described by Thurneysen (1946: 320): “So too,
an adverb or adverbial phrase used in periphrasis with is or ni, which
elsewhere is invariably followed by a formally independent clause, is
found with a nasalizing or a leniting relative clause. Examples: is amne
as coir ‘it is thus that it is filling’ Ml. 114a
l … ; with a leniting clause: ni
fris ru chét ‘it is not with reference to it that it has been sung’ Ml. 64a
l 3,
is dò thucad ‘it is for this it has been cited’ Sg. 45b
19.” However, there
seems to be no precise equivalent of this construction in Wb. Scottish
Gaelic also has the lenited forms of the verb after clefted adverbials,
but they are usually introduced (see my description of the facts 1978:
66 — 67, 69 — 70) by a particle ann, literally ‘in it’. Thus, it might be
tempting to assume that a syntactic distinction lost elsewhere survived
in the north and east, even if the means of displaying it changed rather
radically. At the same time, I do unfortunately not possess enough
evidence to be able to tell exactly when the (is) ann construction in
Scottish Gaelic emerged, so that I still prefer (cf. 1978: 75) to exercise
caution in this matter. In any case, the less conservative syntax of Ml.
and Sg. is again worth noting.
If the syntax of Wb. seems more conservative than that of the two
other collections, there is at least one case where Wb. seems to have
been more systematic in carrying out an innovation than the two other
collections òf glosses. This will appear if the paradigm of tech ‘house’ in
the singular is studied:
NAV tech, teg
G tige, taige
D tig, taig
The forms with non-palatalised t(a)- are obvious innovations, as,
e.g., the Greek cognate tégos will reveal instantly. Their origin10
relatively unimportant to the present argument, but their distribution
it not. Thus, I have no nominative form with ta-, the genitive tige once
in Wb. (7C
9) and taige once in Sg. (66a
19); I have two examples of the
dative tig from Ml. (57c
7; 120d
2) and three of taig from Wb. (9b
30 Anders Ahlqvisí
9; 33a
6). Thus, as Wagner (1983: 104) points out, it “is a fact that
the Würzburg Glosses contain (exclusively) the dative form taig …,
and the St. Gall Glosses the genitive form idal-taigae .. a situation
which suggests strongly that the scribes of the Glosses in question were
Northern Scotti”. In the modern language, the pattern of distribution
is quite clear (Wagner 1958:147,1969:195): Scottish Gaelic (including
Rathlin, about which see further my observations [1978: 68 — 69] on
the cleft adverbial feature) and Manx (i.e., Eastern Gaelic, as defined
by Jackson 1951: 78) have generalised the forms in ta-, whereas
Modern Irish dialects have retained as palatalised t{-e-, -/-) in all cases,
except in Ulster Irish, where both are found, as one might reasonably
expect in a border area. Accordingly, the pattern of distribution of the
dative singular11
of this word in the glosses could be interpreted as
another pointer to the northernness of Wb., in that one might argue
that it represents the locale (paradigmatic and geographic) where the
feature in question originated, before being generalised in northern
and eliminated in southern dialects.
In what precedes, some references have been made to the Wb.
glosses being rather older than the two other collections, At the same
time, heed must be taken of McCone’s (1985: 104) warning that:
the basic chronology is of very doubtful validity, depending as it
does upon entirely a priori assumptions about the relationship
between a disturbingly small number of texts. How, for instance, can
we be sure that the somewhat more evolved language of Ml. in
relation to Wb. can be linked directly to chronology? Since the Ml.
glossator was manifestly more careless in some matters of orthography than his Wb. counterpart (see GO I [Thurneysen 1946:]
4 — 5), a similar carelessness rather than a later date may have been
responsible for his greater proneness to admit later or substandard
forms. In short, we cannot show beyond reasonable doubt that the
bulk of Ml. was written after the bulk of Wb.12
If these arguments are accepted, we may yet be forced to understand
the above features in a new light, namely as functions of possibly
dialectal rather than necessarily chronological differentation. In such a
scenario, Wb. would clearly belong to a more northern locale than Ml.
and Sg.
Unfortunately, there seems to be little very tangible non-linguistic
evidence about the provenance of the glosses, with one exception, if
be brought to light for this claim (Byrne 1984: xix): “Very
Dialect parochialism: the Scottish vowel length rule 31
probably written at Castledermot was the St Gall Priscian, probably in
the year 845.” If and when substantiated, this would not be in conflict
with the above scenario; even then, much would remain to be
The tentative nature of much of the above is worth insisting upon at
this stage, as is the essential unity of the Irish and Scottish Gaelic
speech community until fairly recent times. For this, I would still, on
the whole, follow Jackson’s (1951: 78 — 80) view, according to which it
would appear:
that the modern Irish dialects did not really begin to come into
existence before the thirteenth or fourteenth century at the earliest;
until that time we must suppose Irish to have been a homogeneous
language throughout the island except for such small local differences and incipient nuances of dialect as are bound to occur over
such a comparatively large area. Considering the intimate connexion between Ireland and the Highlands all through the Middle Ages,
and the fact that the two countries were culturally one until the
seventeenth century, there seems, then, a priori no strong reason why
in the tenth century, five centuries after the Dalriadic colonization of
Scotland, Argyllshire Gaelic, for instance, and Manx should have
been any further removed from Ulster Irish than Ulster Irish was
itself from that of Munster; that is to say, at this period little if at all.
Indeed, all that has been discussed in this paper may well be
described as “incipient nuances of dialect”. At the same time, the fact
that they are rather finely gradated dialect features do not make them
any less interesting; perhaps the contrary may yet turn out to be the
Allusion has already been made above to the likely existence of
different registers in the glosses. If we go outside them, much more
material can be found. In this paper, I shall only mention one case,
which concerns the matter of “archaic” Old Irish syntax. It has
generally been assumed by many14
but by no means all15
scholars that
the SOV word order found in some Old Irish sources preserved in late
manuscripts is a genuine survival of a pattern that once was the normal
unmarked one in the sentence. In these studies, much has been made of
the archaic Indo-European nature of the texts. However, evidence is
now becoming available to show that much of the material we now
have was in fact written by authors who used this as a kind of
Kunstsprache. As Breatnach (1984: 459) has put it, “the fact that we
32 Anders Ahlqvist
have evidence for” this “style being practiced in the eighth century, and
used to translate Latin, has important implications for the dating and
nature of other texts in ‘Archaic Irish’.”
Also, the philological problems involved must not be forgotten. For
instance, F. Kelly (1976), in his pioneering edition of Recension Β of
Audacht Morainn, has adhered16
to editorial principles that permit him
to restore much of the language to the standard of the earlier glosses.
Given the sometimes rather chaotic nature of Middle Irish spelling,
this procedure is not without benefits as far as orthography and
phonology are concerned. From a morphosyntactic point of view,
however, similar freedoms can have rather serious consequences. I
have dealt with one, to my mind, not unimportant case of this nature in
a contribution (Ahlqvist 1980: 109)17
to the proceedings of a meeting
belonging to another series devoted to historical linguistics; I have also
published (Ahlqvist 1984) a fresh edition of this text myself: this I have
based on the principle of adhering to manuscript readings as closely as
I then felt to be possible, even when the spelling, as often is the case, can
be interpreted as exhibiting either an archaism or a Middle Irishism.
Since then, the pendulum has swung further. Thus, when quoting some
legal passages in the article already referred to, Breatnach (1984: 445
n. 3) has “compared the text of CIH [= Binchy 1978] with a
photograph of [BM Ms.] Nero A vii”, making his “punctuation, worddivision and capitalisation differ from both CIH and the MS.
Otherwise the text is as it is in the MS and suggested emendations are
given separately”. This may well be the best procedure to follow until
such time as Thurneysen’s hope (as referred to by Binchy 1972:
37 — 38) is realised, according to which “it will be possible to compile a
separate grammar of archaic Old Irish (going well beyond the language
of the Glosses) if and when the archaic stratum of the Laws, the oldest
poetry, and the saga-‘rhetorics’ have been fully investigated.”
In the meantime, the very nature of this future grammar will
obviously be much affected by the editorial labours that precede it,
even to the extent of determining whether it will deal only with
“archaic” Old Irish or attempt to supersede Thurneysen’s work. If
editors continue to restore texts to a standard not far removed from
that of the glosses, the grammar of older and/or (?) higher-register Old
Irish will obviously be rather close to that of Thurneysen’s (1946)
standard, so that McCone’s (1985: 104) warnings cannot be ignored;
according to these, texts
Dialect parochialism: the Scottish vowel length rule 33
of arguably Old Irish provenance that survived only in considerably
later manuscripts are liable to be evaluated and dated with reference
to such a framework, laid on this distinctly creaky bed of Procrustes
and hacked or ‘normalized’ to the form judged appropriate.
Although such normalization can be justified as a purely pedagogical device aimed at sparing learners the need to grapple simultaneously with innovatory forms and the monstrous complexities of
standard Old Irish, it can hardly be defended as a proper editorial
procedure in its own right. … Even more seriously, attempts to
normalize texts surviving in later manuscripts to an ideal standard
based upon Wb. and Ml. are bound to founder upon the fact, amply
enough documented in the foregoing, that these selfsame contemporary sources fail to conform fully to such a standard and
actually contain features of precisely the ‘Middle Irish’ type liable to
be emended out of texts by over-zealous editors.
In this particular connection, it seems appropriate to stress that Old
Irish scholarship has much to learn from what is happening in the
philology of other languages. In a most interesting discussion of
certain grammatical notions in the American Indian language Fox,
Goddard (1984: 285; cf. 1973) issues this timely warning:
if we are ever to understand the subtle patterns, or even the basic
patterns, of the use of proximates and obviatives (and they surely
differ from language to language), we will have to rely upon texts in
which these features have not been tampered with by well-meaning
editors. It is completely unacceptable — a basic inexcusable error of
scientific method — to normalize texts in accordance with our surely
incomplete understanding of Algonquian grammar. Other Algonquianists … have emended proximates to obviatives, and the
defence always seems to be the same (though not always stated):
only clear errors have been emended. I hope this paper shows the
danger of this approach. Don’t do it.
Of course, any problems that an editor perceives in a text can be
dealt with as extensively as desired in textual notes, and where an
emendation is inescapable the apparatus criticus should make the
facts explicit. It is certainly a good sign that the use of these well
established editorial practices is gaining in popularity in the editing
of American Indian texts.
Is it too much to hope that the same devices will, reasonably soon, be
similarly gaining18
in popularity among Old Irish scholars? In this
34 Anders Ahlqvist
respect, we have something to learn from scholars working on Welsh
Accordingly, it may be important to realise that the importance of
the “prestigious forms of Old Irish”20
is largely a function of modern
scholars’ knowledge about the language that was codified in
Thurneysen’s (1946) Grammar. Excellent as this is, it must (as I have
hinted above) be superseded one day by a more up-to-date account,
just as it had itself taken the place of an earlier German edition (1909)
of the same work, which in its turn had replaced the pioneering and
fundamental works of Zeuß and Ebel (1853;
21871). When the new
work appears, in the by no means very near future, it will, to a very
fundamental degree, be influenced by the quality of philological work
involved not merely in editing and reediting Old Irish texts, but also in
appraising the value of their evidence towards depicting the reality of
Old Irish language, as spoken and written in its natural habitat in early
mediaeval Ireland. In other words, scholars now need to find the
answers to this not altogether simple set of questions: who wrote what
where and when?
If they will have succeeded in these aims, the grammar will, or so it
seems to me after this very modest enquiry, include at least a full
chapter devoted to an inventory of what it so far has not been possible
to describe as anything more substantial than “nuances of dialect”, in
our still rather monolithic conception of Classical Old Irish.
* I am most grateful to Jacek Fisiak and all my other Polish hosts for having me at
their most wonderful conference, thus providing the occasion for this paper and for
the very valuable exchange of ideas with other participants, among which many but
especially Anthonij Dees, Ives Goddard, Marie-Louise Liebe-Harkort, and Herbert
Pilch contributed materially to the final shape of this paper.
Also, I owe much to Gearóid Mac Eoin, Nicole Müller, Máirtín Ó Briain and
Donncha Ó hAodha, who discussed it with me in Galway when our Department’s
Research Seminar met on 11 March 1986. Furthermore, I wish to thank Liam
Breatnach and Heinrich Wagner for advice on certain points. Finally, it will of
course be understood that I alone am responsible for any errors and heresies.
1. The standard edition is that of Stokes and Strachan (1901 : xxiii – xxv; 499 – 712). Of
the two other hands, the earlier “prima manus” is given as Wb. I and another one
(somewhat “later” than the main one) is referred to as Wb. II.
2. See the Introduction (xiv—xxi, and esp. xviii) to the standard edition (Stokes—
Strachan 1901: 7 — 483) and Thurneysen’s (1946: 5) comments.
Dialect parochialism: the Scottish vowel length rule 35
3. See Thurneysen (1946: 6) and Byrne (1984: xix) but note that the Introduction
(Stokes—Strachan 1903: xix) to the standard edition (49 — 224) is rather more
cautious on this point.
4. For the most important recent addition, see Bachellery (1964).
5. Even if Irish now has discarded it and pronounces /ifirlc/.
6. In Modern Irish: An fear ar thug mè an leabhar dò; note that dò is literally ‘to him’
and see further Ó Buachalla (1983: 69).
7. As Donncha Ó hAodha has pointed out to me, there is at least one example of this
construction in the legal language: cf. Binchy (1978: 2224.21).
8. A similar position regarding the place where the literary standard established itself
was taken by Gearóid Mac Eoin in a still unpublished paper (“The standardisation
of Old Irish”) given at a colloquium (cf. Ni Chatháin — Richter 1984) held at
University College, Dublin, in May 1981.
9. I have discussed all this at greater length elsewhere (1987: 326-327, 338, 340).
10. Thurneysen (1946: 216) explains that the “forms with a have probably been
influenced by maige, maig, from mag ‘plain’.” This does not convince me.
11. The genitive forms are not very useful, not least because the Sg. one is in a compound
ind idaltaiga ‘illius fani’, in which the non-palatalised t(a)- could, but need not result
(Thurneysen 1946: 98 — 100) from assimilation with the preceding -al.
12. I note in passing that this would dispose rather neatly of McCone’s own previously
expressed strong conviction (1980:17 — 27), based on relative chronology, in favour
of an analogical explanation for the genesis of the Old Irish nasalising relative
clause; I have dealt with this elsewhere (1983: 10—12 [see esp. 12 n. 16] and 1987:
13. This statement is found in the preface to a very useful introduction to Irish
penmanship and palaeography, where argumentation to sustain it might have been
out of place.
14. Cf. Greene (1977), Ahlqvist (1980), and Breatnach’s (1984: 459 n. 1) bibliography.
15. Especially Wagner (1967,1977). In this connection, I might add that my own view
still remains (1980: 109-110) that I agree with Mac Coisdealbha (1976: 314) that
some, but only some, of the SOV material is artificial.
16. E.g., when he (F. Kelly 1976: xxi) feels that it is justified to “restore the in -o
of /’-stems and «-stems, though there is no direct MS. evidence for doing so. It is
however normal in Wb. and almost invariable in archaic sources”. As I have pointed
out (1984:163), its prehistory is somewhat uncertain, and so is its precise status even
in the glosses themselves, as McCone (1985: 87) has shown.
17. See also my edition (Ahlqvist 1984: 156.46) and further remarks in Ahlqvist (1985:
142 n.43), as against F.Kelly’s (1976: 6.46, 29) treatment.
18. It follows that I do not quite agree with Korolyov (1983: 40) when he states that in
“modern editions”, the “reconstructed form should be based on the state of the
Würzburg glosses, preserving archaisms of older strata if such should occur
therein”, even if I feel that he is quite correct in stressing that “we must face the fact
that exact reconstruction is unattainable”.
19. For instance, note that even in a pedagogically orientated text like Thomson’s (1961)
Branwen it is stated (xlvii) that the “orthographical forms of the manuscript sources
are preserved throughout”.
20. Borrowing a phrase used by Ó Buachalla (1985: 5; cf. also 1982: 429) in a timely
warning against over-emphasizing the importance of such materials.
36 Anders Ahlqvist
Ahlqvist, Anders
1978 “On preposed adverbials”, Scottish Gaelic Studies 13/1: 66-79.
1980 “On word-order in Irish”, Papersfrom the 4th International Conference on
Historical Linguistics, ed. by Elizabeth Closs Traugott et al. (Amsterdam:
John Benjamins) 107 — 113.
1983 “Two notes on relative marking in Old Irish”, Celtica 15: 10 — 12.
1984 “Le testament de Morann”, Etudes celtiques 21: 151 —170.
1985 “The relative endings of the Old Irish simple verb”, Ériu 36: 136 — 142.
1987 “The syntax of relative marking in Old Irish”, Folia Linguistica Histórica
6/2: 323-346.
1987 a “An Irish text on the letters of the alphabet”, Studies in honour of René
Derolez, ed. by Anne-Marie Simon-Vandenbergen (Ghent: Dept. of
English and Old Germanic Philology), 3 — 16.
Bachellery, Edouard
1964 “Les gloses irlandaises du manuscrit Paris Latin 10290”, Etudes celtiques
11/1: 100-130.
Binchy, Daniel A.
1972 “The so-called rhetorics of Irish saga”, in: Indo-Celtica. Gedächtnisschrift
für Alf Sommerfeit, ed. by Herbert Pilch and Joachim Thurow (München:
Max Hueber), 29-38.
1978 Corpus Iuris Hibernici I — VI (Dublin: Institute for Advanced Studies).
Breatnach, Liam
1984 “Canon law and secular law in Early Ireland: The significance of Bretha
Nemed”, Peritia 3: 439-459.
Byrne, Francis John
1982 “Cadessin”, Ériu 33: 167 -169.
1984 “Introduction”, The Irish Hand, by Timothy O’Neill (Mountrath: The
Dolmen Press), xi — xxviii.
1913—1976 Dictionary of the Irish Language (Dublin: Royal Irish Academy).
Goddard, Ives
1973 “Philological approaches to the study of North American Indian
languages: documents and documentation”, Current Trends in Linguistics, ed. by Thomas A. Sebeok (The Hague: Mouton) X, 1: 727 — 745.
1984 “The obviative in Fox narrative discourse”, Papers of the Fifteenth
Algonquian Conference, ed. by William Cowan (Ottawa: Carleton University), 273-286.
Greene, David
1974 “The growth of palatalization in Irish”, Transactions of the Philological
Society 1973: 127-136.
1977 “Archaic Irish”, Indogermanisch und Keltisch, ed. by Karl Horst Schmidt
(Wiesbaden: Dr. L. Reichert Verlag), 11-33.
Hessen, Hans
1914 “Ein Fall von Dissimilation im Altirischen”, [Kuhn’s] Zeitschrift für
vergleichende Sprachwissenschaft 46: 1—2.
Dialect parochialism: the Scottish vowel length rule 37
Jackson, Kenneth Hurlstone
1951 “Common Gaelic: the evolution of the Goedelic languages”, Proceedings
of the British Academy 37: 71-97.
Kelly, Fergus
1976 Audacht Morainn (Dublin: Institute for Advanced Studies).
Kelly, Patricia
1982 “Dialekte im Altirischen?”, Sprachwissenschaft in Innsbruck, ed. by
Wolfgang Meid, Hermann Ölberg and Hans Schmeja (Innsbruck:
Beiträge zur Kulturwissenschaft, Sonderheft 50), 85 — 89.
Korolyov, Andrey A.
1986 “A plea for the textus reconstructus”, Proceedings of the Seventh
international Congress of Celtic Studies, ed. by D. Ellis Evans, J.G.
Griffith and E. M. Jope (Oxford: Oxbow Books), 285.
Mac Coisdealbha, Pádraig
1976 The syntax of the sentence in Old Irish (Bonn: Sprachwissenschaftliches
McCone, Kim
1980 “The nasalizing relative clause with object antecedent in the glosses”, Ériu
31: 10-27.
1985 “The Würzburg and Milan glosses: our earliest sources of’Middle Irish’
Ériu 36: 84 -106.
Me Manus, Damian
1983 “A chronology of the Latin loan-words in Early Irish”, Ériu 34: 21 — 71.
Ni Chatháin, Próinséas—Michael Richter
1984 Ireland and Europe. The Early Church (Stuttgart: Klett-Cotta).
Ó Buachalla, Breandán
1977 “Ni and Cha in Ulster Irish”, Ériu 28: 92 -141.
1982 “Scribal practice, philology and historical linguistics”, Papers from the 5th
International Conference on Historical Linguistics ed. by Anders Ahlqvist
(Amsterdam: John Benjamins), 425—432.
1983 “The prepositional relative clause in SE-Ulster Irish”, Celtica 15: 69 — 77.
1985 “The /-future in Modern Irish: a reassessment”, Proceedings of the Royal
Irish Academy 85 C 1: 1 – 36.
Ó Dochartaigh, Cathair
1976 “‘Cha’ and ‘Ni’ in the Irish of Ulster”, Eigse 16: 317 – 336.
O’Rahilly, Cecile
1976 Tain Bò Cúailnge. Recension / (Dublin: Institute for Advanced Studies).
O’Rahilly, Thomas F.
1932 P1972]. Irish dialects past and present (Dublin: Browne and Nolan
pDublin Institute for Advanced Studies]).
Stokes, Whitley—John Strachan
1901-1903 pl975]. Thesaurus Palaohibernicus I —II (Cambridge: University Press
PDublin Institute for Advances Studies]).
Thomson, Derrick
1961 Branwen Uerch Lyr (Dublin: Institute for Advanced Studies).
Thurneysen, Rudolf
1909 Handbuch des Alt-Irischen I. Grammatik (Heidelberg: Winter).
1946 A grammar of Old Irish (Dublin: Institute for Advanced Studies).
38 Anders Ahlqvist
Wagner, Heinrich
1958—1969 Linguistic Atlas and Survey of Irish Dialects I — IV (Dublin: Institute for
Advanced Studies).
1967 “Zur unregelmäßigen Wortstellung in der altirischen Alliterationsdichtung” s. Beiträge Pokorny, ed. by Wolfgang Meid (Innsbruck: Beiträge
zur Kulturwissenschaft), 289 — 314.
1977 “Wortstellung im Keltischen und Indogermanischen”, Indogermanisch
und Keltisch, ed. by Karl Horst Schmidt (Wiesbaden: Dr. Ludwig
Reichert Verlag) 204-235.
1983 “Studies in the history of Gaelic dialects. Part 1”, Zeitschrift für celtische
Philologie 39: 96-116.
1986 “Iarfhocal ar Ni agus Cha sa Ghaeilge [A postscript to Ni and Cha in
Irish]”, Féilscríbhinn Thomàis de Bhaldraithe [The Tomás de Bhaldraithe
Festschrift], ed. by Seosamh Watson (Dublin: Coiste Fhéilscríbhinn
Thomàis de Bhaldraithe), 1-10.
Zeuß, Johann Caspar
1853 Grammatica Celtica (Leipzig: Weidmann).
Zeuß, Johann Caspar — Hermann Ebel
1871 ”•Grammatica Celtica (Berlin: Weidmann)

Dialects Splintering into Languages

I think what I wrote about the Irish language and its precedents likely having dialects and still do when it comes to using certain colour terms (and anything else to some extent, really). If true, this might explain why most dialects of Irish use gorm to mostly mean blue and dark skin whilst a few Irish dialects as well as Scottish Gaelic have it include green vegetation. (There’s a 1768 dictionary stating that ‘gorm’ means both ‘blue’ and ‘black person’.)

This might not be unique to the Irish and Scottish Gaelic languages as a similar phenomenon likely occurred to the Romance languages. Given they all descend from Vulgar/Slangy Latin, this makes the most sense in how and why Italian, Catalan and Romansh preferred domina to mean woman (donna, dona and duonna) whilst French, Occitan and Neapolitan stuck to femina (femme, femena and femna).

Likewise Portuguese and Spanish clung onto mulier (mulher, mujer). Spanish, Portuguese and Italian use casa to mean house but French has it mean hut (though it might not be unique in here since I think a few others like Wallon and Normand might use something similar to the French maison). Since French and Italian actually have lexical similarity of almost 90%, so there’s bound to be a number of common cognates.

The ones I recall are tisotto/tissu, fratello/frere, sorella/soeur, coccinella/cocchinelle, matina/matin (Irish has madin), chien/cane, blond/biondo and singe/scimmia (Milanese Lombard likely has simia/scimia and if I were to give the Irish equivalent, it’s simaog). Let’s not also forget that there are even French words spelt similarly to that of their Italian counterparts (loup/lupo and the afformented matin/matina).

However there are cases where Italian words have rough equivalents in Portuguese, French in Spanish as well as Italian in Spanish and French in Portuguese. The Italian/Spanish words for birthday are compleanno/cumpleaño. The French/Portuguese equivalents are anniversaire/aniversário. The French/Portuguese words for cresent are croissant/crescendo but Italian/Spanish use mezza-luna/media-luna.

As for the striped hyena, in French and Spanish it’s hyène rayée/hiena rayada but in Portuguese and Italian it’s hiena listrada/iena striata. Not to mention Catalan, Portuguese and Spanish use hermano/germano/ermao and bruxo/brujo/bruixot and the like. Both French and Italian use frere/fratello and the like but French uses sorcier whilst Italian uses stregone (though arguably both start with s).

If there’s a dialect continuum in some contemporary Romance languages, the same thing should be said of Vulgar Latin given this is how many Romance languages began.

Blue, Cyan and Green in Irish

Often, almost always undisputedly so in Irish though it also means black person. Though the additional mean of green (as in dark green) might be dialectal though that could also be the case of Irish’s very own precedents, especially when it came to separating into Irish and Scottish Gaelic respectively.

Whilst some dictionaries and texts give the additional meaning of dark green, a 1768 dictionary makes ‘gorm’ restricted to both focal blue and black people. Same with a few other studies. (Keep in mind Mr Patrick Dinneen may’ve come from a different part of Ireland where gorm can also mean green or something.)

Now if you want me to be honest, I actually tend to think of blue as a dark colour and since gorm’s also used to mean black person then gorm might as well be a murky colour in here. (Though this might also be shared with other languages to some extent. Russian has a separate word for focal blue in siniy and goluboy’s not only used to refer to cyan but is also a colour term in its own right.)

Glas–Cyan, Light Green, Vegetation Green
If I’m not mistaken, the word ‘glas’ is used to refer to both vegetation green and sheep grey as well as blue (most especially in Welsh and Breton where it is the focal word for blue). So much so that it’s unmistakably the ‘grue’ word in the Celtic languages (as proven in one of my excerpts). Whilst some Irish speakers use ‘glas’ to mean green in general, I tend to regard it as both plant green and focal turquoise/cyan.

I wouldn’t refer to leaves and grass as ‘uaine’ but glas. I’d even refer to cyan/turquoise items as glas. (Likewise, my own take on blue as well as its Irish counterpart’s almost always restricted to focal/electric blue, blue eyes and dark blue.) I might even regard Irish glas as more or less equivalent to Russian goluboy but in the sense of being a proper colour term for something in-between blue and green.

Not that Irish and even Russian lack grue tendencies but I can go on arguing similar things can be said of English where in my opinion hasn’t had a proper colour term for turquoise until recently whereas Irish already does. (Likewise, Turkmen does in fact have a separate word for blue in mawy/mavy with kok being blue-green/cyan.)

Uaine–Focal/Bright/Artifical Green
Like I said, I actually think/consider Irish to be one of those languages that have a separate colour term referring to turquoise/cyan/blue-green. In fact, it’s right up with Russian, Turkmen and possibly Scottish Gaelic to some extent when it comes to having a definitive word for blue-green from the start. If gorm is blue, glas is cyan (in my opinion) then uaine is focal green.

Uaine’s the green of artificial items like clothing, lights, bright green, eyes and dyes. This is practically how I regard uaine as with glas used for vegetation green, wool grey and cyan items. Funnily enough, in one study it shows that Korean also has two separate colour-terms for green. One is focal green or green proper, one referred to yellow-green. Again, Irish might not be unique in this regard.

Whether if it’s one of those languages with a built-in word for cyan (right up with Russian and Turkmen to some extent) or right up with Korean in having two separate colour-terms referring to certain shades of green.

Note that whilst much of it’s my own opinion, the odd fact that Irish has a separate colour-term for what’s used to encompass both vegetation-green and wool grey as well as a counterpart in Welsh meaning blue makes me think Irish already has a word for ‘blue-green’.

Flesh and Word: Reading Bodies in Old Norse-Icelandic and Early Irish Literature (Excerpt)

Also very prominent in the visualisation of the hero are the symmetrically (!)
distributed four dimples on his cheeks. These are coloured buide, úane, gorm
and corcra (‘yellow’, ‘green’, ‘blue’ and ‘purple’).¹²³ The dimples evoke a sense
of balance on the face, a crucial point given the importance of symmetry McManus proposes in relation to concepts of beauty.¹²⁴ Moreover, their multi-coloured
appearance is also worth noting. This implies that it is not an ordinary, natural,
physical body which is described but rather one exhibiting literary conventions
of expressing beauty. Stark colour contrasts are observable on other noble heroes’ faces in early Irish saga literature also. In TBDD the noble warrior Conall
Cernach – who is introduced as [a]s caime di laechaib Hérenn – is said to feature
two differently coloured cheeks: [g]ilithir sneachta indala grúadh dó, breicdergithir sían a ngrúad n-aile.¹²⁵ In addition to his cheeks, Conall also exhibits a striking colour contrast in his eyes. This supports the previously stated assumption
that multi-coloured facial features and the contrast the colours evoke are an expression of beauty in many early Irish texts.¹²⁶

In terms of imagining Cú Chulainn’s face as a whole it is likely that, even
given that colour terms are a challenging area in early Irish scholarship, the
terms used here do not paint a fair hero: his hair is probably two thirds dark
(dond (‘dark’) and cróderg (‘blood-red’)), with the remaining third presumably
fair like a crown of red gold (mind órbuide).¹²⁷ Furthermore, on his face the
only openly bright colour is buide (‘yellow’). Úaine (‘green’) is somewhat prob-
 All translations of colour terms provided here are referring to the online version of the DIL, (accessed on ..).
 McManus, ‘Good-looking and Irresistible’, p. .
 Quoted from Eichhorn-Mulligan, ‘Politics of Anatomy’, p. ; ‘The handsomest of the heroes of Ériu’; ‘As bright as snow one cheek, as speckled red as foxglove his other.’
 Mulligan mentions the possibility that these descriptions might have had a parallel in reality in the use of make-up or hair-dye and also refers to the codes regarding the colour of
clothes. While this is of course possible, the present study focuses solely on the literary
realm. See Eichhorn-Mulligan, ‘Politics of Anatomy’, p. .
 For an introduction to the topic of colours see Heidi Lazar-Meyn, ‘Colour Terms in Táin Bó
Cúailnge’, in Ulidia: Proceedings of the First International Conference on the Ulster Cycle of Tales.
Belfast and Emain Macha – April , ed. J. P. Mallory and Gerard Stockman (Belfast,
), pp. –.
92 2 Speak for Yourself! Expressive Mediality and the Self
lematic, as according to Stefan Zimmer the term covers various shades of
green¹²⁸ but here most likely expresses a vibrant colour. Corcra is rendered as
‘crimson’ or ‘purple’. Gorm is has a semantic range of ‘blue, deep blue’ and is
apparently often opposed to glass ‘green’, or ‘grue’ (‘green-blue’), a distinction
giving gorm a connotation of ‘dark, black’ also. Cú Chulainn may therefore not
be particularly fair in appearance but rather incorporates rich, dark and saturated colours and maximum colour-contrast.¹²⁹