List of figures
1.1
A hypothetical tree-diagram
1.2
An alternative model for western and southern Indo-European (Garrett, 1999, p. 152)
1.3
A symbolic representation of a Polylectal Internally Dynamic Competence
1.4
The Flux dynamism
2.1
The Kawapanan linguistic area
3.2
A hypothetical Kawapanan family tree
3.3
Shawi vowels
3.4
Vowel ellipses for /i ɘ o a/ in the F1/F2 plane. Each vowel contains 24 scatters (4 words × 2 repetitions × 3 equidistant measurements taken from the steady-state portion). Raw formant values were converted to bark and sigma ellipses were superimposed (number of sigmas = 2) in Praat (Boersma & Weenink, 2015).
3.5
Oscillographic and spectrographic representation of kɘma ‘you’. In this case, no [h] is found.
3.6
Oscillographic and spectrographic representation of kɘkɘn ‘heav’. The friction found in the coda of the first syllable has been marked.
3.7
Semantic scope of the prefix a- (based on Shibatani & Pardeshi, 2002, p. 90,101)
3.8
Sociative causation semantic scope (Shibatani & Pardeshi, 2002, p. 101)
4.1
An example of forms grouped into cognate sets from 4 Austronesian languages (from https://abvd.shh.mpg.de/austronesian/research.php, top left), which are translated into a binary format (bottom left). The right side shows a possible historical tree linking the different forms, with the birth and death of cognate set 2 shown.
4.2
How aligned forms are converted into binary features (left), so that it is possible to model the innovation and replacement of particular phonemes (right). The tree structure is hypothetical, not the suggested best historical analysis for this case.
4.3
Locations of languages in the data.
4.4
A densitree of the posterior distribution.
4.5
The maximum clade credibility tree. Labels are coloured according to Glottolog language family. Numbers indicate the proportion of trees in the sample that include the given split. All nodes in collapsed sections of the tree have posterior probabilities of less than 80%.
4.6
A tanglegram comparison between the tree for Chapacuran languages from this study (left) and from Birchall, Dunn, and Greenhill (2016) (right)
4.7
The Kawapanan-Puelche pronominal clade
4.8
Geographical distribution of the clade
4.12
Comparison between the tree derived from pronouns and trees derived from genetic data. Left: Pemberton et al. (2013); right: Reich et al. (2012).
5.1
Subject Raising, from SOV to OVS (L. M. Rojas-Berscia, ms.) t
1 in this case refers to ‘first tense’ (Seuren, 2018a)
5.2
The vernaculars of the North-East
5.3
Balsapuerto Inciso Aserrado Pottery and Balsapuerto-Inciso Aserrado, Aplicado, Pintado Pottery (Rivas Panduro, 2003)
5.4
Left-branching vs. right-branching (Seuren, 2018a, p. 86)
5.5
Predicate cluster: Arawak a- prefix and the Shawi verb ka’-, for the sentence wa’an a-ka’nin tata atuari ‘The apu makes dad eat chicken meat.’, undergoing predicate raising (PR) and branching directionality change.