User:Erisceres/McCone's Relative Chronology

A summary of Goidelic developments chapter by chapter of Kim McCone's Towards a relative chronology of ancient and medieval Celtic sound change (1996).

Chapter One: The Phonology and Orthography of the Attested Celtic Languages

I.1 Introduction

• Irish Ogam inscriptions date to the 5th and 6th c. C.E.
• The Irish manuscript record emerged in the 7th c. C.E.
• Church literacy provided the model for vernacular writing.
• Ogam is almost certainly based on the Roman alphabet.
• Manuscript orthography from the 5th to 12th c. C.E. was based on British Latin pronunciation.
• After the 12th c. C.E., the orthography underwent experimental changes.

I.5 Ogam Irish

I.5.1 The Origins of Ogam

Cultural Influences

• The earliest attested method of writing Irish is Ogam.
• It is written on a line formed by the edge of a stone.
• It consists of:
  • one to five notches (a, o, u, e, i);
  • one to five horizontal strokes to the right (b, l, f, s, n);
  • one to five horizontal strokes to the left (h, d, t, c, q);
  • one to five diagonal strokes across (m, g, ng, z, r).
• It is almost certainly based on the Roman alphabet.
• Its twenty signs in four groups of five obviously had a numeric basis.
• It was probably devised in the 4th CE.
• There was probably already an established Christian community in the south of Ireland at the time.
• The Latin-based literacy of Christianity possibly influenced its creation.
• Irish colonies were being established in Wales in the 4th CE, providing another possible link to the learning of Latin.
• Archaeological evidence shows a material link between Ireland and Roman-based communities at the time.
• Cultural contact between these communities would have helped provide the environment and influence for its creation.

Innovations

• The alphabetic characters substituted for the Ogam symbols were assigned during the much later manuscript tradition.
• The omission of Latin p, and phonetic pairings of d/t and c/q, show innovation with a practical focus on Primitive Irish phonology.
• These innovations help show the phonemic structure of Irish of the 4th CE.
• The later substitution of alphabetic characters shows deviation from the original phonemic values.

Phonemic Considerations

• Bilingual inscriptions from Wales show the transcription of v instead of the substituted f assigned during the manuscript tradition, considering /f/ was unlikely to have arisen from /w/ during the 4th c. C.E (V.2.2).
• The later reanalysis was probably due to the historical change found in the initial sound of the name of the letter:
  • fern “alder” < *wernā.
• The manuscript tradition did seem to recognise a written distinction between c and q even though the latter was lost in speech by the 6th CE (IV.3.4).
• There was an obvious phonetic connection between the initials of the names for the letters but a lack of phonemic distinction between velar and labiovelar counterparts:
  • coll “hazel” (< *kollo- < *koslo-) for c;
  • cert “apple-tree” (< *kʷerto- < *kʷerxto- < *kʷerkʷ-to- < *perkʷ-; II.1.5b) for q.
• Since the 6th CE merger of /kʷ/ with /k/, a second letter was needed for this written distinction and q was coincidentally chosen for historical /kʷ/.
• Like f, the signs h, ng and z seem unlikely to have been devised to represent the sounds ascribed to them in the manuscript tradition as /h/, /ŋ/ and /z/ were hardly distinct phonemes in 4th or 5th CE Irish.
• Their assignment seems to have been a Latin-based cosmetic choice with their name initials being a contributing factor.
• Their true values may have been /j/, /ɡʷ/ and /sᵗ/ respectively, or something similar, but this cannot be directly tested due to their lack of attestation on available inscriptions.

Phonemic Developments

• For as long as /kʷ/ existed in Irish, so too did /ɡʷ/.
• They first dissimilated to /k/ and /ɡ/ respectively before a following rounded back vowel (i.e. /u(ː)/ or /o(ː)/).
  • Consider the attested dego(s) (which predates apocope) rather than *dengo(s) “flame”, “fire” < *degw-i-s.
  • The ⟨q⟩ in qunacanos is a hypercorrection since the distinction with ⟨c⟩ became lost in this environment before lowering occurred.
  • Cases like tria maqa "(of the) three sons" and ineqaglas show that this dissimilation was only restricted to back vowels which were rounded.
  • The partially Latinised Wroxeter Stone inscription cvnorix macvs maqvi coline "Cunorīx son of Maqqos Colinī" shows how /kʷ/ dissimilated to /k/ first when followed by (rounded) back vowels before it became lost later before front vowels (and all other positions).
• They then rounded a following /a/ and /i/ to /o/ and /u/ respectively (IV.3.4).
• Both finally merged with /k/ and /ɡ/ respectively in the 6th CE (II.1.2) after both raising and lowering (IV.2.1b/c and IV.3.4).