Plenary panel: Comparability

Thursday 14 December

9.00am Plenary Panel: Comparability

[(Session Chair Masha Koptjevskaja-Tamm)] featuring Rachel Nordlinger (University of Melbourne), Jean-Christophe Verstraete (University of Leuven) and Hilary Chappell (L'Ecole des Hautes Etudes en Sciences Sociales-Centre de Recherches Linguistiques sur l'Asie Orientale)

Panel: Greville G. Corbett, Kim Sterelny, Lindell Bromham, and Ulrike Zeshan 

Panel chair: Nick Evans 

Comparability is a central issue in linguistic typology. We presuppose comparability of characterisation across different languages whether we are asking whether a phenomenon is universal or not (e.g., noun phrases or the noun-verb distinction), looking at implicational statements, (say: 'a language will always have as many oral vowel phonemes as nasal ones'), plotting distributions of features in a database like WALS or Glottobank, setting out definitions that can be used in grammatical descriptions, or formulating evolutionary hypotheses about changes from one typological state to another. There have been major debates on this topic within linguistic typology in recent years, but we should not forget that the comparability problem is confined to the study of the language: in biology, anthropology, archaeology and many other fields of endeavour, comparable problems arise. 

This panel will seek to frame problems of comparability in a broader perspective by juxtaposing two linguists (Corbett and Zeshan) with two non-linguists: a philosopher of cultural evolution (Sterelny) and an evolutionary biologist (Bromham). Each will make a short presentation (15 minutes) on their own view of the problem, to be followed by general discussion. 


9.00 - 9.15 Why Fahrenheit Should Have Been a Linguist 

Greville G. Corbett, Surrey Morphology Group 

Fahrenheit was the first to produce two thermometers which, under identical conditions, produced identical readings. In linguistics, given the same phenomenon, we need to produce the same readings more often. The prerequisite for comparison is measurement, and if we take measurement more seriously, we can hope for the virtuous circle of more accurate measurement, which leads to more insightful accounts, which require more accurate measurement, which leads … Of course, a thermometer targets temperature only. If we refine our linguistic scales (criteria), each targeting one phenomenon, we can tease apart combinations of factors (see, for instance, Nikolaeva 2013 on finiteness). This matters since the clustering we observe may be accidental or significant, a point stressed in both Canonical Typology and Multivariate Typology. Given a criterion, we push it to the extreme (as Kelvin did for temperature) and ask whether instances of the extreme actually occur or not (either result is of interest). 

As an example, if we pull apart the possible criteria for reported speech, we find a wide range of possibilities, as Evans (2013) shows. Interestingly, natural languages vary considerably, but each avails itself of few of the possibilities. A large part of the theoretical space is not occupied. Conversely, the extreme instance of inflection classes, as defined by the combination of the extreme values of the canonical criteria, would appear to be highly unlikely on functional grounds. And yet, it is indeed found, in Burmeso (Donohue 2001, discussed in Corbett 2009). 

Once we measure carefully, we are no longer confined to labelling label items as just “hot” or “cold” (as if the world were that simple); rather we can explore their finer-grained nature. Variability and empirical uncertainty become easier to characterize, and can now be incorporated into our analyses, rather than being factored out of them (see Round forthcoming for helpful discussion). As we measure more carefully, additional tools become available, and we can enter mainstream (social) science (Bickel 2015). We do not need exceptional devices for linguistics. Rather we measure the variability of the length of vowels, or the range of the genitive case value, using carefully defined criteria (Corbett 2012); we do so similarly when 

comparing the idiolects of two Russian speakers, when comparing older/younger speakers, Russian speakers with Polish speakers, and Polish speakers with Archi or Tamil speakers. 

It is hard to imagine science without consistent measurement. We should follow Fahrenheit’s lead. 


Bickel, Balthasar. 2015. Distributional typology: statistical inquiries into the dynamics of linguistic diversity. In: Bernd Heine & Heiko Narrog (eds) The Oxford Handbook of Linguistic Analysis, 2nd edition, 901-923. Oxford: Oxford University Press. 

Corbett, Greville G. 2009. Canonical inflectional classes. In: Fabio Montermini, Gilles Boyé and Jesse Tseng (eds) Selected Proceedings of the 6th Décembrettes: Morphology in Bordeaux, 1-11. Somerville, MA: Cascadilla Proceedings Project. Available at: 

Corbett, Greville G. 2012. Features. Cambridge: Cambridge University Press. 

9.15-9.20 Questions and discussion 

9.20 - 9.35 Holism, Atomism and Natural Classification 

Kim Sterelny, Centre for Moral, Social and Political Theory, Australian National University 

One of the main debates about theories of cultural evolution is about the extent to which a community’s culture is a single integrated system, or whether (conversely) it can be seen as a set of independent traits; traits which can change independently of one another, be borrowed into neighbouring cultures, or be found independently in other cultures. Theories of cultural evolution are committed to seeing cultures as ensembles of relatively independent traits, but that view faces considerable challenges. Consider, for example, the role of norms in human social worlds. In western cultures, there is a reasonably robust distinction between moral norms, religious norms, social norms (of politeness and the like) and norms of disgust. To what extent can those distinctions be drawn in other cultures, with their very different social practices, belief systems, and institutional frameworks. A theory of the cultural evolution of moral norms has to be able to identify moral norms, and specify the important ways in which they differ one from another, in different cultural systems. Similar problems arise, for example, for those interested in the origins and social role of religion. Probably there is something somewhat like religion in all known human cultures. But is not obvious that there is a single phenomenon, expressed very differently, in different cultures, or whether there is just a rough and not very informative family resemblance between very different practices embedded in very different cultural systems. This talk discusses these issues, which pose a structurally parallel issue to their applications in analyses of language. Evolutionary approaches to language, likewise, are committed to seeing languages as to a considerable degree ensembles of traits that can change independently of one another, and which can be found with different partners in other languages. 

9.35-9.40 Questions and discussion 

9.40 - 9.55 Comparability in biology: Tracing evolutionary change 

Lindell Bromham, Macroevolution and Macroecology Group, Research School of Biology, Australian National University 

Tracing paths of evolutionary change over time requires us to consider how shared features have been modified in different species since they shared a common ancestor, and this means 

we have to be able to establish comparability between traits. I will illustrate the importance of comparability in evolutionary biology using one of the more obscure yet fascinating examples of Charles Darwin’s scientific researches, his two volume taxonomic treatise on barnacles. Darwin faced the challenge of how to explain the evolutionary trajectory of unique and highly modified traits that appear to have no equivalents in related taxa. He did this by tracing the developmental paths of these unique traits, looking for variation in these strange adaptations between individuals, and comparing them across species that varied in their degree of modification from the ancestor. By using careful observation to establish comparability, even in such an incomparable animal as the barnacle, he could reconstruct plausible evolutionary explanations for even the most bizarrely modified traits, such as the invention of cement to stick them to surfaces include rocks, boats and whales, hermaphroditism and the presence of parasitic males. Nowadays, scientists increasingly rely on DNA evidence to trace evolutionary paths, which bring both advantages and challenges in establishing comparability. 

9.55 - 10.00 Questions and discussion 

10.00 - 10.15 Comparability between signed and spoken languages 

Ulrike Zeshan, University of Central Lancashire, Preston, UK 

Over the past 15 years, sign language typology has developed as a separate sub-discipline of linguistic typology. However, systematic comparisons between signed and spoken languages on a larger scale have yet to be undertaken. Informed by recent efforts towards conceptualising such a cross-modal typology (e.g. Pfau and Steinbach, 2006, Zeshan et al, 2013), this presentation discusses some of the associated challenges and categorisations. 

Sign languages being natural and fully developed human languages, we may expect, as a starting hypothesis, that the two language modalities fulfil largely equivalent communicative functions for their user communities. However, the encoding of communicative functions in terms of linguistic structures and constructions is partly modality-dependent in ways that are not well understood at present. 

It is therefore helpful to start by considering a typology of comparability such as discussed in Zeshan and Palfreyman (2017). In this framework, relative modality effects occur where a certain cross-modally identified structure, while attested in both modalities, is markedly more frequent in one modality than the other. For instance, non-decimal numeral systems seem to be rarer in sign languages than in spoken languages. While this is of interest, the issue of comparability between modalities poses itself most forcefully in the case of absolute modality effects. 

Absolute modality effects involve structures that are only attested in one of the language modalities. In order to argue that this is the case, however, one must make sure that the right standard of comparison is chosen. For example, it is well known that sign languages make use of three-dimensional space for purposes other than mere articulation. In Alipur Sign Language from southern India, the numerals 100, 1,000, and 100,000 are formed by making a successively larger movement in three-dimensional space, while the other aspects of sign formation remain the same (Zeshan et al, 2013). We may argue that such a spatially expressed morphology a priori has no equivalent in spoken languages, and therefore this is automatically identified as an absolute modality effect. However, it is also possible to argue that we should look for the closest comparable phenomenon in the spoken modality that would correspond to enlarged space in the signed modality, for example vowel length. In this case, it is a matter of empirical evidence whether an absolute modality effect can be verified or not. 

Other issues of comparability have to do with the degree to which sign languages have a higher degree of “iconicity”, i.e. non-arbitrary relationships between the form and the meaning 

of signs. For example, many sign languages have numerals whose form mimics a written symbol, often resulting in formational particularities such as handshapes not used elsewhere in the language. Should these be excluded from comparison with spoken languages on the basis of being peripheral to the language, or should they be considered particularly interesting challenges to comparability between modalities? 

In the above examples, we may try to abstract away from the physical form of sound-based words and visual-gestural signs, but this is not always legitimate. Clearly, when comparing the phonological/sub-lexical systems of spoken and signed languages, we must look closely at the physical modality. For instance, in how far are the movement component and the hand configuration in signs similar to the vowels and consonants in speech (Brentari 2002)? Linguistic typology has not properly caught up with such comparisons. 

In this presentation, the aim is to identify and exemplify such issues of comparability, which need to be addressed in any systematic approach to cross-modal typology. 


Brentari, Diane (2002). Modality differences in sign language phonology and morphophonemics. In: Meier, Richard, Kearsy Cormier and David Quinto-Pozos, eds.: Modality and structure in signed and spoken languages, pp. 35-64. Cambridge: CUP. 

Pfau, Roland and Markus Steinbach (2006). Pluralization in sign and in speech: A cross-modal typological study. Linguistic Typology 10.2 (2006): 135-182. 

Zeshan, Ulrike, Cesar Ernesto Escobedo Delgado, Hasan Dikyuva, Sibaji Panda, and Connie De Vos (2013). Cardinal numerals in rural sign languages: Approaching cross-modal typology. Linguistic Typology 17: 357–396. 

Zeshan, Ulrike and Nick Palfreyman (2017). Sign Language Typology. In: The Cambridge Handbook of Linguistic Typology, edited by A. Y. Aikhenvald and R. M. W. Dixon, pp.178-216. Cambridge: CUP. 

10.15 - 10.20 Questions and discussion 

10.20 - 10.30 General discussion

  • Australian Government
  • The University of Queensland
  • Australian National University
  • The University of Melbourne
  • Western Sydney University