How to gloss a manuscript: the view from Milan (Bobbio)

1
How to gloss a manuscript the view from Milan
(Bobbio)

• Aaron Griffith
• University of Vienna

2
Introduction I

• Fact There is extensive Old Irish glossing of
  Latin texts.
• This fact presents an opportunity
• Each language can inform our understanding of the
  other.
• This fact also presents a challenge
• How do we determine what is real Old Irish?
• How do we determine what is real
  (Hiberno-)Latin?

3
Introduction II

• Some recent work focuses on glossing as a
  phenomenon in its own right
• Jacopo Bisagni (Galway) examining code-switching
  between Latin and Old Irish.
• Alderik Blom (Oxford) examining Psalm glossing
  in Northwestern Europe (Germanic and Celtic)

4
My intended contribution here

• My intention here is to
• present various techniques used to gloss Latin in
  the Milan MS (Bibliotheca Ambrosiana C. 301
  infra).
• compare (when relevant) these techniques to
  native grammatical use.
• I hope to offer
• rules the scribes employed when glossing.
• some thoughts on how they thought when glossing.

5
Preliminaries the Milan glosses

• The Manuscript
• from the early to mid 9th century.
• a psalm commentary together with several
  prefaces.
• The commentary
• Julian of Eclanums (c. 386 c. 455) translation
  of Theodore of Mopsuestias (c. 350 428) Greek
  commentary (to Ps. 16 v. 10).
• summary of Julians translation of Theodore (to
  Ps. 16 v. 11).
• The Glosses
• text is fairly heavily glossed in both Latin and
  Old Irish.

6
Some general observations on glosses

• Gloss types
• Translational / Clarificational Glosses
• single word
• short phrase or sentence
• Interpretative / Explanatory Glosses
• generally somewhat longer (at least a clause or
  sentence)
• On interpretative / explanatory glossing
• grammatical
• theological (see McNamara, The Psalms in the
  Early Irish Church. Sheffield (2000).)
• very frequently a historical interpretation of
  the psalms
• sometimes concerned with variant textual readings

7
Glossing techniques to discuss (see Thes I xxi)

• Various single word / case glosses
• e.g. Latin ablative by OIr. prep. ó
• e.g. Lat. alligo (allego) by OIr. conrig /
  asindet
• Latin Participles
• Present active
• Future and perfect passive
• Latin Infinitives
• indirect discourse
• other contexts
• Relatives
• Gender of items not mentioned explicitly in Old
  Irish (e.g. Adjectives)
• Note Treatment of quotations not discussed here
  (see L. Breatnach, On the citation of words and
  a use of the neuter article in Old Irish Ériu
  41, 1990)

8
Various single word / case glosses

• OIr. ó as catch-all gloss for Latin ablative
• 55b8 INDUANTUR CONFUSSIONE8 horuccu by
  disgrace
• 48a5 qua oratione infirmitatis suæ tempore sit
  Ezechias ussurus5 honerberad biuth
  which he would use
• cf. regular (?) OIr. construction 86d12 .i.
  sechip ed arabera biuth i.e.
  whatever it is that a man consumes
• Latin alligo bind ( allego adduce, allege)
• glossed conrig binds in 21b7-8, 23c12
• glossed asindet relates in 23c12, 101a3, 111c5
• glossed adfét relates in 118d10
• see Griffith Varia I Ériu 59, 2009 (153-154)
  for brief discussion

9
Latin Participles I Present active Participles

• glossed with a relative form
• 26b9 psalmum stupentis9 uoce conclusit dicendo
  mothaigedar which is amazed (cf. 96b12, 102b11,
  122a12, 140b5)
• glossed with aN when
• 2d3 quia igitur nuper cum Ebreo disputans3
  arrucestaigser frissinnebride when you (sg)
  disputed with the Hebew (cf. 14d8, 15d10,
  17b16)

10
Latin Participles II Future active (gerundive)

• Usually glossed with copula (usually in past
  subj.) and verbal of nec.
• 79c1 profanamque prosperitatem emitandam1 non
  putetis bed n intamaltai that it
  should be imitated
• 64c3 quanta etiam perficiat miranda3
  innahi ata adamraigthi
  the things that are to be admired
• Minor patterns (with lase / aN when, prep. doL,
  coL full verb, vn.)
• 15d7 non discutiendi7 anambet ecailsi
  when they will not be to be examined (also
  61b1, 63c3, 72b13)
• 53c5 quia ad gustandum5 prouocauerat .i.
  dumlassacht etarcni d? i.e. to taste the
  knowledge of God (also 63a10)
• 54d14 ad interficiendum mé14 celerabant
  coetardamdibitisse in order that they might
  destroy me (also 55b2, 60c6)
• 62a19 ad inpetrandi19 autem facilitatem
  loichtho of obtaining (also 62c11)

11
Latin Participles III Past passive participle

• usually glossed with OIr. past passive
  participle
• 14c13 in malorum fuga uedeatur exposita13
  latharde expounded
• 18c14 temporis quo Abisolón inuasso14
  ambanindrisse when it was invaded

12
Latin infinitives (according to Thes I xxi)

• indirect discourse glossed by an indicative with
  nasalizing relative
• 36d10 etiam per id uoluit adprobare quod Deo
  dicit persequente suos aduersarios interisse10
  asindbathatar that they died (also 25c15)
• other infinitives glossed by subjunctive with
  nasalizing relative
• 30b12 uult dauid autem indicare12 quod eo qui
  in templo habitet defensore utatur
  infé that
  he might point out (also 15a10, 20a9)

13
Indirect discourse in Old Irish

• general case introduced by a simple nasalizing
  relative (GOI p 318 503(g))
• 24d25 asberat immurgu heritic asned dechur
  tabadar leg. tadbadar isindísin heretics,
  however, say that this is the difference that is
  shown therein (cf. 57c4)
• 49b13 .i. durumenar romsa dia i.e. I thought I
  was a god
• special case introduced by nasalizing
  subjunctive when required by subord. clause (GOI
  p 330 518 (e), p 333 520, 2)
• 125c2 .i. asrubart dia hi recht ón arasechitis
  athimnae i.e. that is, God had said in the Law
  that they should follow His commandments

14
OIr. glosses of Latin infinitives in indirect
discourse

• Thes assumes indicative is regular. Basically
  correct
• 25c15 nam Dominus in Euangelio quæ inter
  principia psalmi dicta sunt præsentis sibi
  competere15 demonstrat immindaircet that they
  are appropriate
• 36d10 etiam per id uoluit adprobare quod Deo
  dicit persequente suos aduersarios interisse10
  asindbathatar that they died (also
  54c21, 54c24, 60a3)
• However, they note
• 14d16 cui etiam beatitudinem credit15 rite
  competere16
• 15 erbaid he entrusts
• 16 immandairi that it might be appropriate
• Probably to be explained through the occasional
  use of the subjunctive in such clauses, as noted
  by Thurneysen.

15
Other Latin infinitives glossed by Old Irish

• These are regularly found with the subjunctive
• 56a4 hominem rationis ussu et tui notitia
  præstare4 fecisti doroscea that he is
  pre-eminent
• 56b39 noli emulari siue mirari39 adnamraigther
  .i. nonetaigther .i. adcosnae son no no carae
  that you admire i.e. that you emulate i.e. that
  is, that you strive after or love
• 62b12 ferre12 uix possum follós that
  I endure (also 25a5, 61a11, 62a12)
• Frequently, the form is ambiguous
• 63c4 cupimus gloriari4 nundan mórthar that we be
  magnified

16
Other Latin infinitives

• Infrequently, the indicative is used
• 24c14 ut uelut degito extento Achitofel
  indicare14 uideatur incoisged he used to
  indicate
• 61a16 ab hís qui in terra sunt eum faciet
  honorari16 arammuinfetar feid huili doini talman
  trissa nadamrae sin i.e. that all men of the
  earth will be honored through that marvel
• The explanation for the indicative in these
  examples is unclear.

17
Old Irish versions of other Latin infinitives I

• Most often such constructions appear with a
  verbal noun
• 27b15 ad cobratsidi cumscugud fercæ dæ dothabairt
  digle taranési they desire the stirring of the
  anger of God to inflict vengeance for them
• 98a4 intan as n accubur linn ní duthabairt do
  neuch when we have a desire to give something to
  some one
• 51a19 arnatomnad nech aepert do som bed necen
  donaib hulib anglanad i.e. that no one might
  suppose him to say that it was necessary for all
  to be purified
• 23a5 ní cumcat aithirgi ndodenum they cannot
  work repentance

18
Old Irish versions of other Latin infinitives
II

• Less often, such constructions take an embedded
  verb
• 21b9 isecen dam són nondages daitsiu it is
  necessary for me that I ask You (sg) for them
• 24d14 ?nic domberthar forceill dintitul a
  testimony about the title may be given
• The verb is subjunctive in these cases. This does
  not help explain the anomalous uses of the
  indicative when glossing the Latin.

19
Relatives and interrogatives I

• Direct cases (nom. / acc.)
• As interrogative
• 34d5 quis dabit ex Sion salutem Israhel?ut
  subaudiatur5 intan asmbeirsom cia dobera íc
  dosión when he says, who will give salvation
  from Zion?
• 35a6 quid nos ad hæc adferemus6?
  cidasindisem what shall we declare?
• As relative
• 14b1 qui templum Dei spoliauit1
  dochoimarraig who has
  stripped
• 20c8 quæ munera suæ bonitatis inpertit8
  fundali which he distributes

20
Relatives and interrogatives II

• Oblique cases in Latin are usually translated
  with Latinisms (GOI p 288 460)
• 16a9 ad quem9 ciaduneuch who is it to whom
  (also 33a9, 93a16)
• 17b23 quomodo23 ciachruth asrobar what is the
  manner in which it can be said
• 23b2 in quibus malis2 cia inolcaib in which
  evils
• Sometimes such examples are corrected to native
  Grammar
• 97a5 cuius temporis facta5
• cisi aimser hiforcomnactar ingnimai ón that is,
  what is the time in which the deeds took place

21
Reference to items not mentioned in the OIr.
(esp. Adjectives)

• Thes I xxi notes 73d14
• unde citum14 et efficax sperat auxilium
  dein swift
• déin (acc. sg. fem. of dían) must refer to OIr.
  fortacht (f a) not Lat. auxilium (n o)
• Thes claims this is found in isolated adjectives.
• Question how regular and wide-spread is this
  phenomenon?

22
Matches between Latin and OIr.

• Adjective reflects the Latin case
• 36b4 admota4 percunctatione roitiu set in
  motion
• 61b13 fictís13 uerbís doilbthib
  feigned
• 80c11 casatís11 insidiís fochrataib
  .i. madachaib shaken, i.e. ineffectual
• 90d6 remisa6 studia laxa lax
• also 31c13, 32b12, 36b9, 38c17, 80b8
• Adjective does not reflect the Latin case
• 19a2 (?) ereptum2 restituas inerchelltae
  the one taken away
• 59d4 omni cura4 abiecta domondae
  worldly

23
Mismatches between Latin and Old Irish I

• Words that show the gender of the (unnamed) OIr.
  word
• 16c11 quando sol reuocatus est per ea spatia quæ
  fuerat emensus11
• ithesidi dorumadirsi it is these that it (fem.)
  had measured
• sol is masc., but OIr. grían is fem. hence si
  it
• 36b9 insperata9 morte subduci
• nephfrescestu unexpected
• mors (fem.), but OIr. bás (nt.), hence dat. sg.
  masc. frescestu
• 63a16 ultor uiuolate leg. uiolatae16 Legis
  
• éillidi corrupted
• lex is fem., but OIr. recht is masc., hence gen.
  sg. masc. éillidi
• also 29c1 and 36b4

24
Mismatches between Latin and Old Irish II

• There are ambiguous cases, where the Latin and
  OIr. have the same gender.
• 50c3 salutem daturus3 an dundaberae when
  You (sg) will give it (salus and ícc both
  feminine) see also 131c7
• Most cases are formally ambiguous
• 132c9 fribulas leg. friuulas9 fabulas
  cuitbedcha ridiculous
• acc. pl. nt. (OIr. scél n o) or fem. (Lat. fabula
  f a) cf. 35b20 and 86b3, where scél glosses
  fabula
• see also 40c9, 42c9, 51b26, 57d16, 58b7, 59b13,
  d4, 60b10, 66d20, 71d3, 74c2, 75a14, 15, 87b13a,
  b, 89c6, 94d26, 27, 94c6, 105d1, 116c4, 118b9,
  121c15, 122b4, 125b5, 125d6, 126a12, 13, d7,
  127a4, 12, 130b9, d6, 130d12, 134d4, d6, 138c16

25
Conclusions from the glossing techniques

• It seems that the scribes frame of reference was
  Old Irish (especially clear from adjectives).
• The scribe matched the OIr. to the Latin syntax
  when possible. Something slightly ungrammatical
  sometimes resulted.
• This scenario fits with work on code-switching as
  discussed by Bisagni (Code-switching in the
  Würzburg glosses? paper presented at
  Bilingualism and text transmission in medieval
  texts Utrecht, 31 May 2013.)

26
Code-switching in the glosses I

• Bisagni shows that in the glosses Irish is most
  usually the matrix language into which Latin is
  embedded (94 opposite in 5).
• He also notes that the Matrix Language in such
  contexts is often the first language (which is
  not necessarily the one with higher prestige) (C.
  Myers-Scotton, 1998 Structural uniformities vs.
  community differences in codeswitching, in R.
  Jacobson (ed.), Codeswitching Worldwide,
  Berlin/New York, 91108.)
• Bisagni observes that Milan has relatively low
  rates of intra-clausal code-switching (2)
  relative to Würzburg (21)

27
Code-switching in the glosses II

• This low figure may simply be due to the nature
  of the Milan glosses, which have a very high rate
  of single word / phrase glosses.
• Perhaps the code-switching in Milan occurs not on
  the page, but rather in the scribes mind, i.e.
  before he writes.
• This seems quite clear in the case of adjectives
  agreeing with their (unexpressed) Old Irish noun,
  but it is also visible in the glossing of
  infinitives and relatives.

28
The take-away

• I have exemplified various techniques used to
  gloss the Latin in the Milan glosses
• The techniques are mostly outlined in Thes I xxi.
• For adjective glossing I have provided a more
  detailed examination, since this is a
  particularly revealing area.
• These techniques seem to confirm a high level of
  functional bilingualism with Old Irish as the
  primary language.
• (P. Russell, What was best of every language
  the early history of the Irish language, in D.
  O Croinin (ed.), A New History of Ireland I.
  Prehistoric and Early Ireland, Oxford 2005, p.
  447)

29

• Thank you!