Affixation Conventionalization Hypothesis

1
Affixation Conventionalization Hypothesis

• An Explanation of Conventional Spellings in Mayan
  Writing
• David F. Mora-Marín
• davidmm_at_unc.edu

2
The Problem Fictitious Vowels

• Basic Mayan root shape CVC
• Basic suffix shapes -VC, -CVC
• Script
• CVC(VC) logograms
• CV syllabograms (very few CVC syllabograms)
• Spelling of CVC roots, CVC-VC or CVC-CVC stems
  typically results in a fictitious vowel

• Some such spellings show a synharmonic
  fictitious vowel C1V1C2-C2(V1) or C1V1-C2 (V1)
• E.g. ka-ka for kahk fire
• Generally considered default
• Some show a disharmonic fictitious vowel
  C1V1C2-C2 (V2) or C1V1-C2 (V2)
• E.g. ba-ki for ba(a)k bone
• Problematized

3
Typology of Proposals

• Knorozov 1967
• Justeson 1978
• Bricker 1989
• Justeson 1989
• Hopkins n.d., 2004
• Houston et al. 1998, 2004
• Lacadena and Wichmann 2004
• Justeson 2000
• Mora-Marín 2001, 2002, 2004
• Kaufman 2003
• Boot 2003, 2004
• Anderson 2004
• A few others

• Orthographically-oriented
• Silent vowels
• Neutral vowels
• Underspellings
• Morphological conditioning
• Affixation conventionalization (typical
  suffixing)
• Phonetically or phonologically-oriented
• Echo vowels (phonetic)
• Assimilation to preceding consonant (labial for
  Ca vs. apical for Ci)
• Diacritical of complexity of preceding vowel
  nucleus
• Most approaches regard synharmony as default

4
Disharmony Hypothesis (DH)

• Houston et al. 1998, 2004
• Claim 1 Synharmonic spellings are default
• Claim 2 Disharmonic spellings represent
• CVC (vowel length)
• CVhC (post-V h)
• CVC (post-V )
• Claim 3 No exceptions
• Claim 4 Disharmonic spellings show vowel length
  was preserved in language of texts

• Claim 5 Late Classic increase of synharmonic
  spellings supports loss of phonemic V vs. V
  distinction
• (Claim 6 Suffixes otherwise underrepresented due
  to disharmony principle are represented
  logographically by morphosyllables)

5
Problems with DH

• First, there are counterexamples
• Second, it is acontextual
• Third, synharmonic spellings are not default
• Phonological conditioning
• Morphological conditioning
• Underrepresentation

• Fourth, disharmonic spellings can be accounted
  for by
• Morphological conditioning
• Underrepresentation
• Fifth, there is no need for morphosyllables if
  simple spelling rules are assumed
• Sixth, it uses spellings of terms with uncertain
  phonological shapes or etymologies

6
How to Test Models

• Problems with DH
• Could other approaches work?
• Test of underrepresentation approach
• Study variant spellings of same root across
  identical morphosyntactic contexts
• Test of typical suffixing approach
• Study identical spellings of same root across
  different morphosyntactic contexts
• Study typical suffixes same roots take in modern
  descendant languages

• Do not assume most synharmonic spellings are
  default
• Check for additional phonological conditions
  leading to synharmonic or disharmonic spellings
• Mora-Marín 2001, 2002/2003
• Kaufman 2003

7
Counterexamples

• yi-cha-ki for y-ihchak(-il) 3sERG-claw(-POSS)
  its claw
• Proto-Mayan iSkaq
• H et al. 1998, 2004 discount yi-cha-ki example
  they argue its reconstruction is uncertain
• Since it is possessed, one could argue that its
  second syllable nucleus could be lengthened

• Vowel lengthening?
• POP y-iskaq
• MAM t-xkyaq
• PQM ERG-ixkaq
• KCH r-ixkaq
• QAN y-isqaq chej casco de caballo
• No vowel lengthening upon possession
• Other potential counterexamples

8
Synharmony Not Default

• Justeson 1989 Obligatory synharmony
• C1VC1 roots ka-k(a) for kahk fire,
  ta-t(a) for tät thick
• CV roots mo-(o) for mo macaw, TE(-e) for
  te tree
• Exceptions are contextually accounted for

• Bricker 2002, Mora-Marín 2002, 2003
• C1VC1 sequences -le-l(e) for -lel
  abstractive, tzu-nu-n(u) for tzu()nu()n
  hummingbird
• CV sequences KABA(-a) for ka()ba()
  name

9
Phonological Conditioning I

• Also other cases of V-deletion upon addition of
  suffixes (and an additional syllable or two)
• chu-ka-j(a) chu-k(u)-ji-ya, suggests chuk-aj
  vs. chuk-j-iy(-a)

• Mora-Marín 2002, 2003
• First Context VCCV sequences
• a-k(a)-ta for ahkt-a(j)-Ø (/ahkot-aj/)
  dance-IVZR-3sABS s/he dance(s/d)
• yo-k(o)-bi-l(i) for y-ok-b-il
  (/y-ok-ib-il/) 3sERG-enter-INSTR-POSS his
  entrance/step (yo-ki-bi also attested, showing
  stem without -il)
• Includes ALL positionals, which typically take
  -CVC suffixes

10
Phonological Conditioning II

• Mora-Marín 2002, 2003
• Second context VC-V sequences
• yi-l(i)-a-ji for y-il-ä 3sERG-see-APPL s/he
  sees/saw it
• More common spelling yi-la-ji (no morpheme
  boundary recognized)
• Also u-PAT(-ta)-(t)i-ji, for u-pat-ij (cf.
  MacLeod 2004)
• Morphographemic spellings

11
Morphological Conditioning

• KIN-ni for kin day, sun
• (u-)KIN-ni-li/le for u-kin-il
  3sPOSS/ERG-day/sun-POSS his/her special day
• KIN-ni-chi for ltkinichgt Sun God
• WAY-ya
• WAY-ya-la for way-al alter ego
• WAY-ya-si for way-as alter ego
• Multiple reinforcement

12
Consonant Deletion

• Mathews and Justeson 1984, Bricker 1989, Justeson
  1989, Zender 1999
• ka-se-wa ka-se for ltcaseugt fifth month
  (Bricker 198945)
• AJAW-le-l(e) AJAW-le for -lel abstractive
  (Bricker 198945)
• u-to-ma u-to for uht-om be.finished-POT/FUT
  will be finished
• C-deletion must be considered before regarding a
  given spelling as conventionalized
• It can lead to seemingly synharmonic or
  disharmonic spellings of roots

13
Disharmonic Spellings Morphological Conditioning
I

• CHAK-ki cha-ki
• CHAK-ki cha-ki for chahk lightning, thunder
• u-CHAK-ki-li u-cha-ki-li for u-chahk-il
  his/her/its lightning/thunder(?)

14
Morphological Conditioning II

• a-na-bi
• a-na-bi for ajnahb(-il) he of the lake
• ya-na-bi-li for y-ajnahb-il his/her
  lake-person
• Naj Tunich mural Aj Nahb(-il) saw it He
  is the Aj Nahb-il of Mr. So-and-So

15
C-deletion Test Variation in Equivalent Contexts
16
Affixation Conventionalization Hypothesis (ACH)

• C1V1C2(-C2V2) or C1V1-C2 (V2) in phonetic
  complementation or phonetic spellings of
  root/word-closing segments the second vowel is
  likely to correspond to the vowel of the most
  common suffix or suffixes that a root may exhibit
  in the texts
• Requires alternative spellings of same root
  across different morphosyntactic contexts (unique
  spellings are useless)
• Can be tested against purely linguistic sources
  (e.g. Kaufman 2003a, 2003b), which can take on
  spellings attested across contexts, as well as
  unique spellings

17
Kaufman 2003a, 2003b

• Kaufman uses primarily linguistic sources, but
  also epigraphic sources
• Some of his data fills in epigraphic gaps in ACH
  approach
• u-ne-na attested as s-nehn-al in Tzotzil
• Some of the gaps in his data are filled in by
  epigraphic data of ACH approach
• CHAK mu-ti-l(a) for red bird
• AJ-to-ka-l(a) for ajtok-al

18
Reanalysis of H. et al.s Data Set

• Disharmonic spellings
• 15/35 spellings accounted for by ACH alone
• 15/35 accounted for by linguistic data (including
  12 not accounted for by ACH)
• 27/34 disharmonic spellings accounted for
• Of remaining 8 spellings 5 are unique or
  invariant, 1 is of questionable etymology, and 2
  are CV-CV spellings of word-final -VC suffixes,
  which are overwhelmingly spelled with
  word-closing Ca signs if discounted, then the
  testable items are all accounted for

• Synharmonic spellings
• 14/39 spellings accounted for by ACH
• 12/39 spellings accounted for by linguistic data
  alone (including 5 not accounted for by ACH)
• 12/39 spellings accounted for by phonological
  conditioning
• Total of 29/39 synharmonic spellings accounted
  for
• Some spellings are unique or decontextualized

19
Conclusions

• ACH has internal control No isolated or unique
  examples admitted
• ACH depends on system-internal evidence only
  (actual attested spellings)
• Underspellings are a fact, and theyre common
  (and attested in colonial Yukatek documents
  written by scribes who applied hieroglyphic
  conventions to the alphabetic medium)
• Other phonetic, phonological accounts are
  unnecessary
• (Does not rely on morphosyllables, only on
  simple, well-attested spelling rules)

• Underspellings
• Disharmonic
• Synharmonic
• Affixation Conventionalization
• Disharmonic
• Synharmonic
• Phonological Contexts (C1VC1, CV, VCCV,
  VC-V)
• Synharmonic Vowel Insertion
• (Polymorphemic Logograms and phonetic complements
  (not MH))

20
Conclusions

• DH does not recognize prevalence of
  underrepresentation
• Does not recognize phonological conditions for
  synharmony
• It is acontextual
• Does not take affixation into account
• Does not take syntax and pragmatics into account

• Its decontextualization of spellings precludes
  testing of other approaches
• Uses spellings of uncertain phonological shapes
  or etymologies
• (Relies on hypothetical existence of
  morphosyllables)
• There are counterexamples