Whorfian hypothesis strong weak work

Of words and trees

If you want to annoy linguists, there are three very simple ways: ask them how many languages ​​they speak, ask them their linguistic opinion on the spelling reform, or ask them if it's true that the Eskimos 50 different words for Have snow. What the three questions, which at first glance seem so different, have to do with each other, is that they put your finger in three big wounds of language research and squeeze really hard.

The question of that Talented linguists aims at the self-image of linguistics practitioners, which says that to do linguistics, you actually only need one language, preferably English (if some syntax fanatics have their way). In a slightly different way, one could also say: mastering just one language has never prevented a linguist from making statements about the development of entire language families, because knowing how to describe languages ​​does not necessarily mean that one speaks them . After all, mathematicians don't have to be good at arithmetic either. But when you ask linguists about it, they don't always like to hear it, especially when they actually belong to those who were brought up in a strictly monolingual manner.

The question of the spelling reform aims at the social relevance of language research. For me personally, it represents the point in time when linguistics failed really grandly for the last time. By staying out of a discussion about the reform as much as possible, she left the field to the literary creators, who then put their personal aesthetic sensibility first, rather than seriously considering the benefit of reforms for those who actually write, to deal with. At the same time, more and more institutes of comparative linguistics began to be closed because the researchers were often unable to clearly convey the contribution their work was making to society. The fact that the historical comparison of languages ​​does have something to say when it comes to the reform of writing systems, which represent systems that have evolved over time, was once again completely ignored - even by the linguists themselves.

The last question concerns the tiresome debate about the Sapir-Whorf hypothesis, which is surrounded by many myths and fake news, that is to say one of the most extensive internal debate within linguisticsHowever, due to its possible scope, it is taken up in many other scientific disciplines and often misused there for internal debates. In its strong form (Whorf 1950), the hypothesis states that speaking affects thinking so directly that we could develop a new form of relativity if only we would finally start doing our science in other languages ​​of the world instead to use the Indo-European languages ​​such as English or German. If Eskimos have 50 different words for snow, that shows (apparently) that their language alone makes them think very differently than we, who are now beginning to forget what snow actually is.

I find it gratuitious to assume that a Hopi who knows only the Hopi language and the cultural ideas of his own society has the same notions, often supposed to be intuitions, of time and space that we have, and that are generally assumed to be universal . In particular, he has no general notion or intuition of TIME as a smooth flowing continuum in which everything in the universe proceeds at an equal rate, out of a future, through a present, into a past; or, in which, to reverse the picture, the observer is being carried in the stream of duration continuously away from a past and into a future. (Whorf 1950: 67)

The Sapir-Whorf hypothesis annoys many linguists (at least me) because it has long been refuted in its naive, strong form, based on the original data that Benjamin Lee Whorf himself used to develop this hypothesis . Whorf’s original argument that there are languages ​​in which timeduration not metaphorically as timelength is recorded, namely on the actual linguistic data Not be detected. It seems - at least that is the current state of research - that linguistic modeling of duration as length, of time as space - is something that seems obvious to all people, no matter what language they speak. Although there is hardly any evidence for the strong Sapir-Whorf hypothesis, it is still repeated, discussed and, above all, unfortunately used as an instrument in non-linguistic circles (for example in debates about the justice of language).

Whether for a weakness Form of the hypothesis that would then say that language unites one conscience Being able to influence ways of thinking and finding evidence is another question that is, however, very difficult to answer in itself. It may well be possible that our thoughts are sometimes shaped in a certain direction by the material with which we can verbalize them. If we like instead of shades of color Light Blue and dark blue two completely different color words, such as goluboj and sin’ij in Russian or celeste and azul in Spanish, it can happen that we develop other thoughts if someone tells us something about Blue cheese that tells in Spanish rather than Dark blue cheese (queso azul) is described,

This does not mean, however, that someone who speaks German cannot tell the difference between light and dark blue just because the difference between the two colors is hardly made in German. Perhaps the strict distinction between two shades of blue in Russian and Spanish leads to a slight increase in attention to these shades, but it is completely unclear how big this effect actually is, and for how many people in the population it really works Strangely enough, if we stick to the blue cheese, it also shows that in Russian one speaks of Light blue cheese (golubye syri) speaks.

Individual languages ​​are neither a template nor a mirror of human thought. They don't necessarily have to direct our thinking and can only provide small clues about how we perceive things. If, for example, a language expresses different concepts such as “hand” and “arm” with the same word, then this can be an indication that hand and arm are in principle not that dissimilar, or that they belong together functionally and often seen by us as a unit. This is the case in Russian, where there is just one word ruka that covers both concepts. This works amazingly well without leading to real misunderstandings because in most cases the context makes it clear what is actually meant.

These kettles or Kolexifications, as we now call the phenomenon (François 2008), are very common in the languages ​​of the world. On the one hand, this has to do with the ambiguity of our words, because no word names just a single concept completely unambiguously; on the other hand, they result from random similarities that have arisen historically. The not accidental Kolexifications on Polysemy are most interesting to the linguists because, if we look at enough data together, they can give us a glimpse into human cognition.

I already presented the procedure I am thinking of in a little more detail in an article from 2018. It is based on collecting colexifications for as many concepts as possible for as many languages ​​as possible, and these as Kolexification network in which the concepts represent nodes in the network, and the connections indicate how often a particular pair of concepts has been colexified, i.e. expressed with the same word, in different languages ​​of the world.

After we updated our Database of Cross-Linguistic Colexifications (CLICS, https://clics.clld.org), we have now succeeded in increasing the data even further, so that we can increase the number of languages ​​whose data has been incorporated into the database from approx. 1200 could double about 2400. In addition, we were also able to further refine the workflows with the help of which we aggregate the data from various sources so that they can now also be easily reproduced (cf. Rzymski et al. 2020).

Even more beautiful than the data itself, however, was a study that colleagues from psychology at the University of North Carolina had initiated with our data, and which we were able to publish recently after more than two years of work (Jackson et al. 2019). In this study, the colexifications of emotion concepts such as “love”, “pity”, “surprise” and “fear” were compared and the resulting network structures in different language families were then statistically compared with one another. The amazing result was that - in contrast to color concepts - the structures of the networks for different language families differ greatly from one another. For example, some language families tend to use the same word for "surprise" and "horror", while other languages ​​use the same word for "pity" and "love".

Not all aspects of the network structure were different. An extended analysis showed that above all the criterion of valence, that is, whether something is perceived as negative or positive, for which networks in all language families played a fairly similar role, and the same could also be true for the degree of excitement (engl. arousal) can be determined.

These results show that the verbalization of emotion concepts is on the one hand very strongly influenced by culture, but on the other hand also seems to be subject to fundamental cognitive aspects that affect all languages ​​equally. What about the results though Not suggests is that those who speak tongues where “compassion” and “love” are spoken of by one word do not know the difference between the two feelings. Because here, too, what I have already said about the colors applies: just because you don't always make a distinction between light and dark blue does not mean that you cannot distinguish between the two shades.

Even if it seems tempting to pull the old Sapir-Whorf hypothesis out of the drawer in connection with our study of emotions and to dust it off carefully, the results still do not provide a single concrete indication that our thinking is shaped by our languages could be. Many factors shape what we think. Language is one aspect among many here. We shouldn't focus so much on Which Language we speak, rather how we speak the language in which we want to express ourselves.


  • François, Alexandre (2008): Semantic maps and the typology of colexification: intertwining polysemous networks across languages. In: Vanhove, Martine (ed.): From polysemy to semantic change. Amsterdam: Benjamin. 163-215.
  • Joshua Conrad Jackson and Joseph Watts and Teague R. Henry and List, Johann-Mattis and Peter J. Mucha and Robert Forkel and Simon J. Greenhill and Kristen Lindquist (2019): Emotion semantics show both cultural variation and universal structure. Science 366.6472. 1517-1522. URL: https://science.sciencemag.org/content/366/6472/1517, DOI: 10.1126 / science.aaw8160
  • Rzymski, Christoph and Tiago Tresoldi and Simon Greenhill and Mei-Shin Wu and Nathanael E. Schweikhard and Maria Koptjevskaja-Tamm and Volker Gast and Timotheus A. Bodt and Abbie Hantgan and Gereon A. Kaiping and Sophie Chang and Yunfan Lai and Natalia Morozova and Heini Arjava and Nataliia Hübler and Ezequiel Koile and Steve Pepper and Mariann Proos and Briana Van Epps and Ingrid Blanco and Carolin Hundt and Sergei Monakhov and Kristina Pianykh and Sallona Ramesh and Russell D. Gray and Robert Forkel and List, Johann-Mattis (2020) : The Database of Cross-Linguistic Colexifications, reproducible analysis of cross-linguistic polysemies. Scientific data 7.13. 1-12. URL: https://www.nature.com/articles/s41597-019-0341-x, DOI: 10.1038 / s41597-019-0341-x
  • Benjamin Lee Whorf (1950): An American Indian Model of the Universe. International Journal of American Linguistics 16.2. 67-72.
You can cite this article as: Johann-Mattis List, "From words to deeds ?," in Of words and trees, 13/01/2020, https://wub.hypotheses.org/1049.
This article was published in Semantics on by Johann-Mattis List. Keywords: CLICS, concepts of emotions, Sapir-Whorf hypothesis.