i6
INTRODUCTION
Glossators II-V seem to have paid more attention to the meaning
of the text than Lucas. Glossator III in particular (Cairbre) clearly
did a lot of work on the text, and corrected many of the main scribe’s
errors, possibly with the aid of another copy of BB (see Introd. p. 6).
In general, the glosses of glossators II-V seem to be their own
composition, based on their own interpretation of the text and their
own knowledge of law. For example, in 23c Cairbre gives bualad na
ces ‘ striking the hives ’ as an example of an illegality which a passerby might do to bees. It seems likely that this is his own suggestion,
not taken from some other source. Similarly in 49c glossator IV
gives Loch Bel Set (perhaps in Co. Tipperary) as an example of an
inaccessible place. As this gloss is only of local relevance, it is likely
to be original. Sometimes, however, the later glossators have drawn
on other sources. Thus in 391 Cairbre’s gloss derge mbortuil is
apparently for indeirge mborbtuile ‘ desertion of violent lust ’ used
to gloss ben deirig a lanamnus ( CIH 233.24 = AL v 456.27). In
52d his gloss a cethair i ncen is taken from the text of BG (see
notes).
Most of the commentary in A is in Lucas’s hand, presumably
transcribed from the same source as the glosses. Sections of commentary have been added from other sources by glossators II and
III (App. 1 (b, f, g, h)). Taken together, the commentary in A
contains a reasonably full precis of the whole text. There is also
commentary in C (App. 2), D (App. 3) and L (App. 4). Much of
the BB commentary in these four MSS clearly goes back to a common
origin. Thus, the summary of §§1-7 in A (App. 1 (a)) is similarly
worded to C (App. 2 (a)) e.g. gidh sochctidhi ita ina riachtanas a
leas isan erich a neineacht , no go null acht in cutruma sin doib uile
A = cid sochaide isin [erich] ata ina riachtanus a leas nocha thabharthar acht an cutruma sin doib uile C. Again, commentary to §49 by
the third A glossator (App. 1 (f, g)) agrees fairly closely with commentary in L (App. 4 (a, b)).
In general, the glosses and commentary to BB derive directly
from the O. Ir. text, and consist merely of paraphrase, explanation
or expansion of the original. Consequently, glosses and commentary
seldom provide additional information about the Irish law of beekeeping (at any period). Nonetheless they have been of considerable
help in our edition. Apart from the occasions where a gloss has
preserved a better reading than the text (discussed above) there are
four main ways in which they can assist:
INTRODUCTION
17
(1) Firstly, the glosses — and to a lesser extent the commentary
— can shed light on linguistic difficulties in the text. If we are right
in our dating of the B glosses and commentary to the 9th century,
this scholiast was working on a text which was approximately 200
years old. The language of the original seems to have caused him
little difficulty, though he v/as clearly unsure of the meaning of the
phrase a/labrig n-ai in § 6. He may also have been unfamiliar with
the word snuaid in §12, as his gloss arna rossnuaidter looks like a
guess (but see notes).
The later glossators, working between about 500 and 900 years
after the composition of BB, still show a fairly good knowledge of
O. Ir. Even forms which are unlike anything in late Middle or early
Modern Irish are often correctly identified. For example, one might
expect a form like §29 rodmbi to baffle a later glossator. However,
the A gloss rogonastar e shows that it was correctly identified as 3
sg. perfect with 3 sg. masc. infixed pronoun. Infixed pronouns are
also correctly interpreted at §18 fodaloing (gl. imuilnges iad ), §50
rosnuccad (gl. rogatad se iat), etc. In §49 the glossator correctly
identified the form doruillet (3 pi. pres. ind. with ro of do-sli 4 earns,
deserves ’) as is clear from his gloss airiltnigid 4 they deserve ’. In
§54 he supplies the correct meaning for -imdich (gl. -didnend
4 defends ’) even though this cpd. does not seem to have been used
after the O. Ir. period (see DIL s. v. imm-dich). Not surprisingly,
however, the later glossators are sometimes wrong. For example,
§39 cain cannot be 2 sg. imperative of canaid 4 sings, chants ’ as
suggested by the gloss aisneidh leat. In §42 rofoiscet cannot be from
ro-saig or saigid as suggested by the A gloss roindsaige siad. Here B's
gloss A. foscugud argues in favour of it being 3 pi. perfective pres,
subj. of fo-scoichi (see notes).
(2) Many of the A glosses are 4 etymological ’ in the Latin
grammatical tradition: a word in the text is broken up into two or
three elements, each of which is explained individually. For example
Lucas provides §4 iar n-aimseraib with the conventional gloss
iarsan re suthain 4 after the long period ’. Here the word aimsir (- er )
4 time ’ is divided into am and sir. Am is taken to be for am(m)
4 time ’ and explained by the word re 4 time, period ’. Sir is taken
to be the adj. sir 4 long ’. Needless to say, this explanation is quite
fanciful: only very occasionally do such etymological glosses agree
with the findings of modern philology. (An example is Cormac s
Glossary, Anecdota iv 86, where the glossator is doubtless correct
i8
INTRODUCTION
in taking disc ‘ a young ewe ’ to consist of the elements ot 4 sheep ’
and sesc 4 dry ’ ; cf. W. hesbin 4 yearling ewe ’ from hesb ‘ dry ’).
4 Etymological ’ glosses can be of assistance where the main text
is doubtful. For example, in §14 the text of A has the puzzling
verbal form inidicoimge , glossed cm/? cuimhges. As cm (c/m) 4 quick ’
is the regular 4 etymological ’ gloss on the preverb it is likely
that the original verbal form began with imid-. We tentatively read
imid-chomba (see notes). As ada 4 lawful ’ is a regular gloss on the
infixed pron. -id-, we emend §12 incuirither to inid-chuirethar (see
notes). The corresponding A glossd has inadacuirither.
However, the reconstruction of the O. Ir. text on the basis of
4 etymological ’ glosses can be complicated by the use of the same
4 etymological * gloss for different words. For example, besgne
4 custom, regulation ’ is normally glossed ba-fis gnae 4 pleasant good
knowledge’, e.g. CIH 210.22; 520.3. Occasionally, however, bes
4 custom, food-rent’ is also glossed ba-fis gnae , e.g. CIH 210.3;
468.10. For this reason we would not be justified in emending §14
a besaib fidbreth to a bescnaib fidbreth in spite of the gloss a ba-[f]is
gnae no aibind breithemhnais na feadh 4 according to pleasant or
fair good knowledge of the judgment of the trees ’ ( aibind is a
direct gloss on gnae). In §27 we have emended in am to ind amsir
on account of the etymological gloss isan re suthain. However, this
emendation is not certain, as re suthain may occasionally have been
used to gloss am as well as amser (though we have found no
examples.)
The practice of providing a word with alternative 4 etymological ’
glosses is common in legal MSS. (Bergin’s witty parody of this
technique is quoted by Dr. Binchy in his 1943 Rhys Memorial
Fecture The Linguistic and Historical Value of the Irish Law Tracts
p. 20). The only example from BB is in §26 where foma is glossed
by Fucas ./. is fo ci mhaith no is foemhta 4 good is his good or it is
accepted ’. Cairbre adds the direct gloss no is rogo 4 or it is a
choice ’.
It should be noted that the use of an 4 etymological ’ gloss does
not necessarily prove that the glossator was unable to understand
the O. Ir. form. Thus the phrase iar n-aimseraib in §4 would have
presented no problem to an Early Mod. Ir. glossator. The gloss
iarsan re suthain is therefore superfluous as an aid to understanding,
and is perhaps best regarded as learned ornamentation. On the
other hand, it is clear that the glossator could not identify tetechtai
INTRODUCTION
19
in §§36, 39, etc. (we take it to be nom. pi. past participle of do-etet
‘ tracks ’). His ‘ etymological 1 gloss tectaid ait no aide ‘ which have
a place or dwelling ’ is therefore an attempt to explain an unfamiliar
form by the Isidorean method of splitting it up into its component
parts. The same technique is used by the glossators to explain the
obscure word allabrig in §6. Glossator II suggests brug ara miad *1
all ara doimne ‘ brag (“land”) for its dignity, and all (“cliff1’)
for its depth
In our translation of BB we mark ‘etymological 1 explanations
in inverted commas. We use hyphens to indicate the component
parts of such glosses, e.g. 49 1 og-ai ‘ perfect one 1 which glosses
dg(a)e ‘ chief (of kindred)’.
(3) The 9th century B glosses and commentary, though not
very extensive, are of some help in elucidating the legal content of
BB and also provide information on other legal matters e.g. 15° (on
trees), 45° (on swearing), 49' (showing the spread of the privileges
of the flaith ‘ lord ’), 52c, 53c (on the classification of livestock).
But even B’s glosses and commentary are sometimes at variance
with the text. His claim in 3U that the value of the tairgille ‘ forepledge 1 given by a beekeeper to his neighbours depends on ‘ the
nearness and contiguity of the [bees’] grazing’ contradicts §2 of
the text (though sec note to §3 i lldge). In I6a he can hardly be
correct in his suggestion that the value of a tree increases in summer
and autumn because of the possibility of swarms alighting on its
branches. In 1 8;i he takes the fintiu griain to refer to the place where
the hives are situated, but this destroys the author’s analogy with
the kin-group. In 41 b he seems to have taken the regulations on the
killing of one bullock by another in a jointly owned herd, and
misapplied them to bees (see notes).
The later glosses and commentary display considerable variation
in their understanding of the text. In some cases, the legal content
of BB is covered quite adequately, e.g. in glossator M’s precis of
§§1-25 (App. I (b)). The same glossator (48 e) detects an inconsistency in §§46-8 of BB. The author differentiates between bees
which have been present for less than a year (§46), and bees which
have been present for more than a year (§47). In §48, however,
he makes no reference to the length of time the bees have been
present. The glossator suggests that by analogy with §§46-7 — the
finder should gel one third of the produce if the bees have been present
for less than a year, and two thirds if present for more than a year.
20
INTRODUCTION
In spite of a few such cases where the later glossators have
spotted a legal difficulty, their glosses are in general not to be
trusted as a source of information on Early Irish law. Errors of
interpretation are frequent in the glosses: for example in 2C the
glossator’s second suggestion that ciapa meit ciapa laiget could
refer to the bees is quite impossible. Similarly, gloss 19b is wrong
in its claim that 4 the meaning of deolaid (“ gratis ”) is that it is not
necessary to cast lots \ In fact, swarms are given gratis to all four
neighbours, even when the casting of lots is necessary (as in the
situation described in §21). In addition to actual errors, many
glosses are irrelevant to the text, such as the reference to the ‘ armed
man on lawful business ’ in 28a, or to 4 levying in the fourth hostile
territory not beyond an arm of the sea ’ in 43c. Equally irrelevant is
the suggestion in 47a that the text refers to swarms which have settled
in sacred trees ( bilida ); in fact it is clear that the nature of the tree
has no bearing on the division of the bees’ produce in this
paragraph.
The glossators frequently make unreal distinctions, not justified
by BB or any other O. Ir. law-tract. For example in §28, it is
unlikely that the author of BB intended any sharp distinction of
meaning between olcc 4 harm ’ and annrecht 4 illegality ’. The first
A glosses, however, take olcc to refer to the killing of bees, and
annrecht to robbing them. In §45 there is no support for glossator
Ill’s suggestion that the forgall is an oath sworn by the neighbours,
whereas the airthech is sworn by the owner of the tree where the
bees settle.
In §§50, 52-3 glossators I and III attempt to distinguish the
meanings of the O. Ir. verbs for stealing fo-tlen, gattaid and fo-coislea
(for which see notes to §50). The distinctions which they make are
unreal, and sometimes contradictory. In 50c and 52b glossator I
takes fo-tlen to refer to stealing 4 some of it ’ (i.e. of the honey) and
glossator III adds 4 some of the honeycombs ’. In 50d and 53c
glossator I takes gattaid to refer to stealing 4 all of it ’. This would
seem to mean 4 stealing all of the honey ’, though glossator III adds
na cesa 4 the hives ’ in 50d, agreeing with the main text rather than
the previous glossator. In 52c glossator 1 explains fo-coislea with
the same gloss he uses for gattaid in 50d and 53c. In 53b, however,
he explains fo-coislea with the gloss he uses for fo-tlen in 50c and
52b. It is obvious, therefore, that the separate meanings which he
assigns to these three verbs are mere guesses.
INTRODUCTION
21
The glossators often provide an example rather than a definition
of a term or phrase in the text. Thus in 3C caechad ‘ blinding ’ is
only one example of cin ‘ crime ’ by a bee (the choice is suggested
by §§30-35). Similarly, in 25a sale is only one of the reasons why
a bee-hive might come under new ownership. Cf., also, 27b, 27c,
46°, 48°, 48d.
The commentary to BB strays a good deal further from legal
reality than the glosses, no doubt because the link with the O. Ir.
text is more tenuous. Some commentary is patently ludicrous, as
for example when the D commentator distinguishes a fourth category
of swarm, i.e. beich aerda ‘ bees in the air ’, in addition to the three
categories ( beich fesa , beich chundtabartacha , beich anfesa ‘ bees of
known, doubtful and unknown origin ’) used by other glossators
and commentators. So far as we can see, no conceivable legal
situation calls for this category of ‘ bees in the air ’. Other commentary provides for extraordinarily remote possibilites. Thus C
commentary (App. 2 (c)) deals with the branch growing from one
man’s land overhanging his neighbour’s land, and even includes
the case where the tree is regarded as a sacred tree by one neighbour
but an ordinary ‘ tree of neighbourhood ’ by the other, and the
branch is cut by a third party!
However, in spite of absurdities and over-elaboration, the legal
content of later commentary cannot be totally written off. Thus the
triad ‘ three areas of a king: a road and unshared land and the
sea ’ (App. 1 (g) = App. 4 (b)) may contain genuine legal tradition.
Commentary in H.3.18 (which we quote in App. 6 p. 190) provides a
convincing explanation of why bee-trespass was regarded as an
offence in early Ir. Law. The long commentary on ‘ injury by bees ’
(CIH 316.37-318.29 = AL iii 432-40) starts with the statement
‘ a hive for blinding, two hives for killing; and the book tells of the
hive for his blinding and does not tell of the two hives for killing.’
The book referred to presumably contained Bechbretha, which
covers injury (especially blinding) by bees in §§27-35, but makes no
reference to death from a bee-sting (see note to 30d). The commentator’s statement is important because it shows that he did not
regard BB as the sole source of Irish bee-law. There is no other
reference in the surviving Irish law-texts to the payment of two hives
in the case of a fatal bee-sting, but it may have been actual legal
practice at some period. For the Welsh law on fatal bee-stings, see
App. 7 p. 205.
OO
INTRODUCTION
(4) Finally, the glosses and commentary provide some scraps of
information on bee-keeping at various periods. It is significant that
they often use different bee-keeping terminology to the main text.
Thus the main text refers to the second swarm as the tanaise, but
the B glossator and L commentator call it the tarbsaithe ‘ bull
swarm \ The third swarm is the meraige ‘ fool ’ in the main text
(= smeraige in A commentary) but iarsciithe ‘ after-swarm ’ in L
commentary. In the main text and in B glosses and commentary,
the hive is referred to as lestar, but in later glosses and commentary
(A, C, L) it is usually ce(i)s.
These differences in terminology suggest that the glossators were
not working on the main text in total isolation from real life, and
were drawing to some extent on their own knowledge of contemporary bee-keeping. The bee-keeping observations of the 9th
century B glossator are particularly interesting. It is clear from his
gloss 29 e that he knew that a honeybee normally dies after stinging
a person. If we are correct in our interpretation of his gloss 25 a (see
notes) he knew of the existence of drones, which he termed tairb
‘ bulls \ On the other hand, he is surely wrong in following the
text’s claim in §5 that a colony’s second year is its ‘ year of
weakness ’ : we have found no support for his suggestion that ‘ it is
in the second year that scarcity and death are most likely to affect
bees ’. In 44e we can see no sense in his statement that ‘ the proper
abode for the bees is beside a bush ’ (see notes).
The first A glosses contribute little information on bee-keeping,
but glossator III, Cairbre, working in the 16th century, makes a few
interesting observations. His gloss 27a implies that the bees of his
day were killed at the extraction of the honey (see note to §27 collud).
His gloss 52a shows that bee-hives were sometimes put out on a
bank or in a thicket, presumably for easier access to a source of
honey. He is the only glossator to refer to the honey-comb, criathar
(50c, 52b, 53b).
From the bee-keeping point of view the most interesting
commentary is L (App. 4 (c)) which implies that a hive may produce
a swarm (‘ first swarm ’) which in turn may produce another swarm
(‘ son ’) which may itself produce a further swarm (4 grandson ’)
— all in the same season. See notes and diagram to §5 bliadain a
sil.
We conclude by drawing attention to three general points about
the glosses and commentary to BB.
INTRODUCTION
23
(1) There is no continuity between the 9th century B glosses
and commentary, and the later glosses and commentary (A, C, D, L).
It is probable that the B glosses are the work of a member (or
members) of a legal family, but not of one with which the legal
families of the later period held continuity of tradition.
(2) The process of accretion of glosses on a law-text is well
illustrated in A, where a word or phrase of BB may be glossed in
one, two, three or (in 52d) four different hands, ranging in date from
the 14th to 16th centuries.
(3) The first A glossator, Lucas, worked in the 14th century,
but we believe that most of his glosses and commentary were
composed somewhat earlier. The dating of legal glosses and
commentary is particularly difficult. This is partly because of the
artificial type of language employed, especially in glosses. Another
difficulty is the scribes’ readiness to alter or modernise forms in
glosses and commentary (in contrast to their usually reverent
treatment of O. Ir. text). There is also a tendency for glossators to
introduce early forms from the text into the glosses, e.g. 50f bite.
Obviously, such forms cannot safely be used for dating purposes.
There is evidence that many law-texts were provided with an
apparatus of glosses and commentary in or around the 12th century.
For example, the version of Coic Conara Fugill in the mid 12th
century MS Rawlinson B 502 ( CIH 2200-3) has fairly extensive
glosses (some interlinear, some incorporated in the text) and
commentary (all of it incorporated in the text). The material
incorporated in the text must be older than the MS. The first A
glosses and commentary to BB , however, do not seem earlier than
the 13th century. We have noted the following dating features (for
which see 6 Cathain ‘ Some studies in the development from Middle
to Modern Irish, based on the Annals of Ulster,’ ZCP xix 1-47).
(1) There is greater use of the past preverb ro (14e, 50f, 52e, 53e
rosuidiged, 29 e rogonastar , 33 a rabatar, 44a, 45 d rainaitcestar, 45b
roelo) than of do (31a docaechadar, 32a dobui , 45b dabadar ), cf. 6
Cathain pp. 14-5. (2) There are 26 examples of independent pronouns as verbal objects, and no infixed pronouns, cf. 6 Cathain
pp. 4-8. (3) There is one example of an - adar ending in place of
the ^-preterite (31a docaechadar ), cf. O Cathain pp. 20-30.
In the A commentary in other hands (App. 1 (b), (f)-(h)) there are
no infixed or independent pronouns. There are two cases of ro
(both in the hand of glossator II) against none of do , and two 3