Evolutionary Linguistics, Alien Languages and Duality of Patterning: What can Slide Whistles tell us about the Evolutionary Emergence of Combinatorial Phonology?
- Lilian Fullerton, University of Edinburgh
- Old College Lorimer Room, University of Edinburgh
Duality of patterning (Hockett 1960, henceforth DoP) describes the existence of two levels of
linguistic structure: combinatorial phonology and compositional semantics. Where combinatorial
phonology is the reuse and recombination of small, meaningless units (e.g. phonemes) into larger,
meaningful units, compositional semantics is the reuse and recombination of meaningful units
(morphemes; words) into constructions where the meaning of the whole can be determined as a
function of the meaning of the parts.
A fundamental question in language evolution research is to understand the origins and
development of DoP. With regards to combinatorial phonology, questions arise as to how and
under what conditions quasi-continuous articulatory spaces (e.g. the vowel space) become
segmented into the productive combinatorial units used in language.
Some insight can be gained from newly-forming natural languages, e.g. Al-Sayyid Bedouin
Sign Language (Israel and Sandler 2010), but these are rare, and experimental manipulation thus
isolating causative factors is difficult. Computational modelling can provide hints and has included
simulations exploring how combinatorial phonology can emerge from holistic coding, typically
through optimization for certain constraints (e.g. Lindblom et al. 1984; Zuidema and de Boer 2009).
Experimental work complements the above methods by enabling precise understanding of how
real speakers may act. Most work on DoP has focused on the evolution of compositional semantics
(e.g. Kirby et al. 2008; Galantucci 2005). It is only recently that these methods are being adapted
to explore combinatorial phonology, where studies have looked at the effects of iterated learning
on its development (e.g. del Guidiche et al. 2010; Verhoef 2010; Verhoef 2011).
The current paper presents a novel experimental paradigm investigating whether combinatorial
structure emerges as a result of pair-wise communication: Participants engage in a game
requesting and giving semantically unrelated novel objects, where the roles of ‘requester’ and
‘giver’ alternate each round. They jointly receive positive feedback if the object requested matches
the object given. Participants are separated and communicate solely through slide whistles, which
mirror the semi-continuous articulatory space. There are no pre-defined signals. Outputted signals
are analysed for combinatoriality, and the way in which the structure of the systems changes over
time is also analysed. Whilst no results are yet available, the prediction is that participants will
progress from a more holistic to a more combinatorial system over the course of the game.
The paradigm presents an opportunity to better understand and test parameters affecting the
degree of combinatorial reuse and thus compliments the growing body of work investigating
computationally, experimentally and with real languages the evolutionary emergence of duality of
patterning.
For ULAB 2011, the background body of contemporary and exciting work in evolutionary
linguistics will be introduced, and the experiment and future potential experiments summarized.
Some preliminary results will be outlined.
There will be opportunity for audience members to play slide whistles.
linguistic structure: combinatorial phonology and compositional semantics. Where combinatorial
phonology is the reuse and recombination of small, meaningless units (e.g. phonemes) into larger,
meaningful units, compositional semantics is the reuse and recombination of meaningful units
(morphemes; words) into constructions where the meaning of the whole can be determined as a
function of the meaning of the parts.
A fundamental question in language evolution research is to understand the origins and
development of DoP. With regards to combinatorial phonology, questions arise as to how and
under what conditions quasi-continuous articulatory spaces (e.g. the vowel space) become
segmented into the productive combinatorial units used in language.
Some insight can be gained from newly-forming natural languages, e.g. Al-Sayyid Bedouin
Sign Language (Israel and Sandler 2010), but these are rare, and experimental manipulation thus
isolating causative factors is difficult. Computational modelling can provide hints and has included
simulations exploring how combinatorial phonology can emerge from holistic coding, typically
through optimization for certain constraints (e.g. Lindblom et al. 1984; Zuidema and de Boer 2009).
Experimental work complements the above methods by enabling precise understanding of how
real speakers may act. Most work on DoP has focused on the evolution of compositional semantics
(e.g. Kirby et al. 2008; Galantucci 2005). It is only recently that these methods are being adapted
to explore combinatorial phonology, where studies have looked at the effects of iterated learning
on its development (e.g. del Guidiche et al. 2010; Verhoef 2010; Verhoef 2011).
The current paper presents a novel experimental paradigm investigating whether combinatorial
structure emerges as a result of pair-wise communication: Participants engage in a game
requesting and giving semantically unrelated novel objects, where the roles of ‘requester’ and
‘giver’ alternate each round. They jointly receive positive feedback if the object requested matches
the object given. Participants are separated and communicate solely through slide whistles, which
mirror the semi-continuous articulatory space. There are no pre-defined signals. Outputted signals
are analysed for combinatoriality, and the way in which the structure of the systems changes over
time is also analysed. Whilst no results are yet available, the prediction is that participants will
progress from a more holistic to a more combinatorial system over the course of the game.
The paradigm presents an opportunity to better understand and test parameters affecting the
degree of combinatorial reuse and thus compliments the growing body of work investigating
computationally, experimentally and with real languages the evolutionary emergence of duality of
patterning.
For ULAB 2011, the background body of contemporary and exciting work in evolutionary
linguistics will be introduced, and the experiment and future potential experiments summarized.
Some preliminary results will be outlined.
There will be opportunity for audience members to play slide whistles.