As we have seen in previous instances of this blog post series, there are many parallels but also many differences between the evolutionary branches of biology and of linguistics. In the following, I will present a comparison of the causes due to which two related inheritable entities (e.g. two words or two genes of different languages or species) may differ from each other, or two unrelated ones resemble each other. The linguistic categories presented here can also be found in List (2016) whereas the biological categories are largely based on Koonin (2005).
Genes that are related to each other are called homologues. There is no well-established linguistic catch-all term for words that are related to each other, but they could be said to have an etymological relation. The most common causes due to which such related entities differ from each other in related languages or species are the processes of regular sound change and random gene mutation, respectively. Both of these happen all the time, and affect all words or genes in the same way (with certain exceptions), and can therefore be considered the default baseline process that does not demand a different name for the type of homology (Gévaudan 2007, 14). Only their absence (as in for example certain iconic words that did not take part in regular sound change, like English peep /piːp/ having retained its Middle English form instead of turning into /paɪp/ (Flaksman 2017, 17)) or irregular cases of sound change caused by frequency effects (Bybee 2017) or analogy, or areas of the genome especially prone to mutation may warrant new terms.
There are also other processes that can be involved in the relation between individual genes or words but they are quite different from each other in the different fields and cannot as easily be equated with each other. For language, there are the various kinds of word formation processes in which a new word is derived based on existing words. The closest counterpart in genetics is gene duplication (List 2014, 40), in which the same gene is included twice, thereby increasing the amount of genes in the genome. As with word formation products, duplicate genes then continue evolving more or less independently from each other but still may have an influence on each other’s future evolution. This is related to their functional similarity in the genome (Koonin 2005, 318, 323), and to processes of analogy in the lexicon.
However, a product of word formation rarely has the same shape as the word it is derived from (the only exception being the process of conversion), whereas duplicated genes initially are the same, only later they may diverge by mutation. Additionally, word formation processes normally involve either more than one word or morpheme as the input (e.g. in compounding or affixation) or at least combine a word with a pattern to derive the new word. The processes in biology most similar to that are fusion and fission, i.e. the merger of two genes and the split of a gene into two, but here the original genes cease to exist, whereas for example compounding does not cause the simplexes to stop being part of the lexicon. These different processes nevertheless share the fact that they challenge the established types of homology (Koonin 2005, 327) and cognacy, respectively.
Xenology is a special type of homology in which a horizontal transmission was involved (Koonin 2005, 315). This happens either in the context of hybridization, when genomes of different species are combined by interspecies reproduction, or, in microbes, by a variety of mechanisms. In the tree model it leads to reticulation. Its obvious linguistic equivalent are borrowings from one language to another (Atkinson and Gray 2005, 513). Hybrids could be likened to creole languages (Gray et al. 2007, 10).
In natural species there are normally only four letters of the genetic alphabet, the four chemicals guanine, cytosine, adenine and thymine. Yet also artificial new letters have been created (Hoshika et al. 2019). Once these molecules undergo reproduction, the replication mechanism might encounter problems interpreting them and treat them like errors (Ledbetter et al. 2018). Similar things happen in language: Since different languages normally have different phoneme systems, the shape of a borrowed word is typically adapted to the phoneme system of the borrowing language (see for example Lin (2008) on the adaptation of English words borrowed into Mandarin). Yet overall linguistic change is much less restricted.
To summarize, in regards to homology, only a few corresponding processes can be found when comparing evolutionary genetics and linguistics, even though the basic ideas are the same.
Other factors of similarity
Common ancestry is not the only possible cause of similarities. Two of the causes most frequently considered are chance resemblances and universal tendencies. Both of these are called homoplasy, making this the complement term to homology as it describes similarity independent of common ancestry.
If languages or species are only very distantly related, similarities between them are usually considered chance resemblances. In languages, words can with reasonable certainty be proven to go back to a common ancestor by searching for patterns of regular sound correspondences between the languages (Lass 1997, 129). Failing to find any regular sound correspondences that would explain the similarity between two words does however not necessarily disprove their relatedness as it can be confounded by processes of analogy and word formation. On the other hand, in some cases actual homoplasies are not purely due to chance, but may result from universal tendencies (List 2016, 120), e.g. high vowels being associated with smallness (Sidhu and Pexman 2018) or words derived from baby-talk having simple syllable structures. As a third possibility, a borrowing can be postulated as the cause.
In a comparable way to linguists putting their trust in regular sound correspondences over surface similarity, biologists tend to trust comparisons on the genetic more than comparisons on the phenetic (morphological) level, the latter of which tend to suffer from increased degrees of homoplasy, as in the very basic example of both birds and bats developing wings. Also in the evolution of species, universal tendencies are at play, meaning the same evolutionary development might occur independently in different species based on universal or environment-specific demands on natural life. Therefore, phenetic similarity is mainly considered in those cases where genetic data is not available (e.g. in fossiles).
But also closely related languages or species may share characteristics independent of any of the causes mentioned thus far even though it is known that their common ancestor did not have them. This is called parallel evolution, or, in linguistics, drift. It means that both descendants changed an ancestral characteristic in a parallel fashion. This is believed to be caused by interdependencies within the system of a given language or genome that cause them to evolve in similar directions independent from each other, combined with universal tendencies (List et al. 2016, 9f). One basic linguistic example for this are processes of word formation. The same derivation, especially in the case of the ubiquitous process of compounding, can easily be formed independently in the same manner in two related languages and it is often difficult to determine whether the word formation in question was already created in the ancestor or only in the daughter languages (or both).
What may result in a pattern in the data similar to homoplasy is the phenomenon of incomplete lineage sorting. Here, a polymorphism (i.e. different members of a population having partially different genes) in the ancestor species is resolved in different directions in daughter species (Nakhleh 2013, 721; List et al. 2016, 6f; Jacques and List 2019, see also the recent post by Jacques) which can mean that more distantly related branches share homologous characteristics that their respective closer relatives are missing.
This has been thus far mostly ignored in computational linguistics even though examples for polymorphisms can also be found in language evolution. A language may have two synonymous words, while each of its two daughter languages loses a different synonym. Or it may have suppletive forms in a paradigm that either end up spreading out onto the whole paradigm and removing the regular word forms, or getting replaced by regular word forms. Unlike in species, the polymorphism occurs not in different organisms but within the same lexicon, potentially within the same utterance. A polymorphism between different mental lexicons of the same population on the other hand is exhibited by sociolects, in which speakers utilize slightly different languages depending on the social context.
If these reasons for similarities can be excluded, the shared innovations may be interpreted to point towards a closer relationship between the descendants sharing them.
Linguistics tends to work with much less data than genetics, so shortcomings of the model that genetics can simply skip over due to the large amount of data they have will come much more to the forefront in linguistics, forcing us to be very clear in what we are actually trying to model. Whether we adopt biological methods in linguistics or develop linguistic models for computational approaches independently, describing the differences and similarities in the way I tried to do in this blogpost can serve as a new perspective to see which aspects may have been neglected.
Atkinson, Quentin D., and Russell D. Gray. 2005. “Curious Parallels and Curious Connections: Phylogenetic Thinking in Biology and Historical Linguistics.” Systematic Biology 54 (4): 513–26. http://www.jstor.org/stable/20061257.
Bybee, Joan L. 2017. “Grammatical and Lexical Factors in Sound Change: A Usage-Based Approach.” Language Variation and Change 29: 273–300.
Flaksman, Maria. 2017. “Iconic Treadmill Hypothesis: The Reasons Behind Continuous Onomatopoeic Coinage.” In Dimensions of Iconicity, edited by Angelika Zirker, Matthias Bauer, Olga Fischer, and Christina Ljungberg, 15–38. John Benjamins Publishing Company. https://doi.org/10.1075/ill.15.02fla.
Gévaudan, Paul. 2007. Typologie des lexikalischen Wandels: Bedeutungswandel, Wortbildung und Entlehnung am Beispiel der romanischen Sprachen. Stauffenburg-Linguistik ; 45. Tübingen: Stauffenburg.
Gray, Russell D., Simon J. Greenhill, and Malcolm D. Ross. 2007. “The Pleasures and Perils of Darwinzing Culture (with Phylogenies).” Biological Theory 2 (4): 360–75.
Hoshika, Shuichi, Nicole A. Leal, Myong-Jung Kim, Myong-Sang Kim, Nilesh B. Karalkar, Hyo-Joong Kim, Alison M. Bates, et al. 2019. “Hachimoji DNA and RNA: A Genetic System with Eight Building Blocks.” Synthetic Biology 363: 884–87.
Jacques, Guillaume, and Johann-Mattis List. 2019. “Save the Trees: Why We Need Tree Models in Linguistic Reconstruction (and When We Should Apply Them).” Journal of Historical Linguistics 9 (1): 128–67. https://doi.org/10.1075/jhl.17008.mat.
Koonin, Eugene V. 2005. “Orthologs, Paralogs, and Evolutionary Genomics.” Annual Review of Genetics 39: 309–38. https://doi.org/10.1146/annurev.genet.39.073003.114725.
Lass, Roger. 1997. Historical Linguistics and Language Change. Cambridge: Cambridge University Press.
Ledbetter, Michael P., Rebekah J. Karadeema, and Floyd E. Romesberg. 2018. “Reprograming the Replisome of a Semisynthetic Organism for the Expansion of the Genetic Alphabet.” Journal of the American Chemical Society 140: 758–65.
Lin, Yen-Hwei. 2008. “Patterned Vowel Variation in Mandarin Loanword Adaptation: Evidence from a Dictionary Corpus.” In Proceedings of the 20th North American Conference on Chinese Linguistics (Naccl-20), edited by Marjorie K. M. Chan and Hana Kang, 1:175–87. Columbus, Ohio.
List, Johann-Mattis. 2014. “Sequence Comparison in Historical Linguistics.” PhD thesis, Heinrich-Heine Universität Düsseldorf.
List, Johann-Mattis. 2016. “Beyond Cognacy: Historical Relations Between Words and Their Implication for Phylogenetic Reconstruction.” Journal of Language Evolution 1 (2): 119–36. https://doi.org/10.1093/jole/lzw006.
List, Johann-Mattis, Jananan Sylvestre Pathmanathan, Philippe Lopez, and Eric Bapteste. 2016. “Unity and Disunity in Evolutionary Sciences: Process-Based Analogies Open Common Research Avenues for Biology and Linguistics.” Biology Direct 11 (39): 1–17.
Nakhleh, Luay. 2013. “Computational Approaches to Species Phylogeny Inference and Gene Tree Reconciliation.” Trends in Ecology and Evolution 28 (12): 719–28.
Sidhu, David M., and Penny M. Pexman. 2018. “Five Mechanisms of Sound Symbolic Association.” Psychon Bull Rev 25: 1619–43.