1 of 14

Creativity & Constraint

Conlanging Beyond the IPA

Logan R. Kearsley

2 of 14

The “Blank Page Effect”

  • Paradox of choice – when you can choose anything, it’s easiest to choose nothing.
  • Easier to work with something that’s already “on the page” than to generate content de novo.
  • Constraints allow us to make choices by solving problems.

3 of 14

Variety in Non-Human Languages

  • Syntax, memory limit effects, discourse structure
    • All neat topics, but for another time

  • Only looking at phonology
    • I work with romanization–you may lack that constraint!
    • “Phonology” includes non-audio modalities!

4 of 14

Alternate Human Modalities

  • Oral/Aural Language
  • Visual Sign Language
  • Tactile Sign Language
  • Writing
  • Whistling
    • More like writing than like sign, but easier to invent!
    • Whistle registers have regular properties related to their “normal” counterparts (Conlangery #152)
    • What if a whistle language existed independently?

5 of 14

Tjugem - A Whistled Language

  • Timbre – sound quality – is out.
  • We have two dimensions to work with: amplitude, and pitch
  • Let’s select a maximally-distinctive subset of the amplitude & pitch patterns that natural whistle registers use.

  • “Vowels” – steady tones, smooth changes, amplitude peaks.
    • Based on three relative frequency bands–high tone, mid tone, and low tone
  • “Consonants” – patterns of alterations to vowels, amplitude troughs.
    • Frequency: below adjacent vowel, above adjacent vowel, or stay the same?
    • Amplitude: change fast, change slow, have a silent gap

6 of 14

Tjugem - A Whistled Language

Consonants:

I C D�A t d n�M k g q�G p b m

Vowels:

HH = i

LL = u

HM = ja

LM = wa

Phonotactics

High-target consonants can’t be adjacent to low-target consonants, because that requires a “long” silence to reconfigure the vocal tract / instrument, or passage through the vowel space, which is another vowel.

HLM = ju

LHM = wi

Schwa = e (H/L/M)

7 of 14

A Canine Phonology

  • Lots of shared mammalian vocal physiology–lips, teeth, tongue, nasal cavity, etc. But…
  • Dogs have a small, but horizontally-arranged, vowel space
    • Not the triangle/trapezoid humans have
    • Entirely outside the human space, so romanization will be purely arbitrary
  • Dogs may not be able to distinguish minimal pairs
    • Phonology based on syllable-level features

8 of 14

A Canine Phonology

  • CV(C) syllables, so every syllable has at least two segments
  • Every feature must alter at least two segments
    • By construction, there will be no segmental minimal pairs!
  • Vowels are front <e> and back <o>
  • Features:
    • +/-back, +/-round, +/-open, +/-stop, +/nasal

9 of 14

A Canine Phonology: Features to Phones

  • -back: front vowel, alveolar/labial consonants
  • +back: back vowel, velar consonants
  • -round: plain onset and plain vowel
  • +round: labialized onset and rounded vowel
  • -open: short vowel and fricative coda consonant
  • +open: long vowel and no coda
  • -stop -nasal: fricative onset and modal voice vowel
  • +stop -nasal: plosive onset and breathy vowel
  • +stop +nasal: nasal onset and nasalized vowel

Syllable count comparable to some human language phonemic inventories. Not a huge number, but we can always add more distinctive features!

Onsets:

f x

t k

n ng

tw kw

Codas:

f x

Vowels recoverable from the consonant string & syllable boundaries!

10 of 14

Fysh A - A Language in Electric Fields

  • Spoken by alien electric “fish”, using electrocyte organs.
  • Based on signalling abilities and limitations seen in real-world electric fish.
  • Three sets of electric organs, of two types with different wave patterns
    • 3 channels, two of which are indistinguishable from each other (so frequency order doesn’t matter) but distinguishable from the third.
  • Amplitude and signal frequency are related–it takes time to build a larger peak charge–so (unlike whistling) each channel only has one dimension.
  • Electrocytes can’t instantaneously transition between discrete frequencies
  • Controlling the start time of a wave train is easier than precisely controlling the stop time.

11 of 14

Fysh A - A Language in Electric Fields

  • “Independent” (syllable-initial) phones: chords of two or three formants
    • Uniquely identifiable by frequency ratios, regardless of transposition.
  • “Dependent” phones: one or two non-fundamental formants
  • Syllables drop formants over time
  • Romanization based on syllables–which formants are dropped, after how much time
    • 1. Simple “Devoiced”: A single sonorant letter.
    • 2. Simple “Voiced”: A sonorant letter followed by a vowel letter.
    • 3. Complex “Devoiced”: A voiceless obstruent letter followed by a vowel letter.
    • 4. Complex “Voiced”: A voiced obstruent letter followed by a vowel letter.
  • Phonotactics:
    • Adjacent segments/syllables in a word must share a formant at their boundaries.

12 of 14

A Cephalopod Phonology

  • No specific cephalopod– just using general features of cephalopod physiology to derive interesting constraints.
  • Actual cephalopods are color blind, but that’s boring–Let’s use their color for communication, not just monochrome patterns! (I.e., this is a constraint I choose to ignore, in favor of others.)
  • Three layers of color-changing cells:
    • Iridophores – iridescent blues and greens
    • Leucophores – white background, can hide or reveal iridophores.
    • Chromatophores – black, red, orange, brown, yellow, in varying saturations
  • Segments come with a foreground and a background color
    • Phonotactic/featural constraint: chromatophore colors cannot be backgrounds, except against black, because they are physically on top!
  • Most colors can be saturated or desaturated, depending on leucophore activation.
  • Allowed combinations of color features make a large number of “vowel”-like segments.
  • Average level of leucophore activation gives us paling/darkening “suprasegmentals”

13 of 14

A Cephalopod Phonology

Static Patterns

  1. Large spots (one per sub-field)
  2. Large rings
  3. Small spots (many per sub-field)
  4. Small rings
  5. Zebra striping
  6. Bands

  • For romanization purposes: onset vs. coda consonants?
  • Lots of neurologically-derived constraints on which static patterns can pair with which dynamic patterns
  • Probably more sensible to make analogies with sign language phonology.

Dynamic Patterns

  1. Large Shimmer
  2. Boundary Shimmer
  3. Fast Flashing
  4. Slow Flashing
  5. Vertical Pulses
  6. Horizontal Pulses
  7. Vertical Rhythmic Waves
  8. Horizontal Rhythmic Waves
  9. Multidirectional Waves

14 of 14

Q&A?