Among the aspects of Serres’ paper that provoked the most animated discussion this morning was his contention, in ‘Information and Thinking’ and elsewhere, that human beings are not the only entities to think. The idea was articulated most powerfully in the snippet of the talk reproduced below:
This thought returned in the final paragraph of the address:
They [material elements] encode, we encode; they count, we count; we speak, they speak.
The thought that exercised some of the panel members in the round-table discussion following the intervention was the way in which ascribing language (and by implication, it seemed to be extrapolated, agency and responsibility) to non-human objects was a metaphor unwarranted by the very clear differences between a human being and (to use Serres’ example from Verne’s cave) a crystal.
This understanding of how Serres is deploying terms like ‘encode’, ‘speak’ and ‘count’ is not the only one available, however, and indeed it is not the one that cuts best with the grain of his arguments. The issue can be clarified by considering the difference between metaphor and metonymy.
The contributions of some panelists seemed to imply that there are only two options available to us: either a ‘flat ontology’ (attributed to Serres) in which there is no way to account for distinctively human responsibility (for the environment, for example), and a human exceptionalism that eschews a flat ontology but legitimates political and social concerns in a way that the flat ontology cannot.
This is not a dichotomy I recognise from my own reading of Serres, and I think that the confusion comes through taking Serres to be speaking metaphorically when in fact he is speaking metonymically.
The (spurious) metaphorical reading of Serres goes something like this: Language simpliciter, language as such, is human language. All other uses of the term ‘language’ are (legitimate or illegitimate) metaphorical extensions of this literal sense of the term. ‘Humans speak’ is not metaphorical; ‘material elements speak’ is. The metaphorical sense here is parasitic on the literal (human) sense; if there is no ‘plain’ sense of what it means to speak, then there can be no metaphorical sense. We start with human language—or human responsibility—as traditionally understood and then every other instance of the term ‘language’ or ‘responsibility’ must be a metaphorical extension of that first, human sense. It’s when we think of Serres in this way that he seems to be painting a Romantic, enchanted, Toy Story world in which stones whisper to each other behind our backs and objects come out to play when we are absent.
But I don’t think that reading of Serres is most faithful to what he is seeking to communicate. We get closer to Serres’ thought when we read him as speaking not metaphorically but metonymically, or more precisely synecdochically (I argue this at greater length in a forthcoming article on Serres in SubStance). In arguing that material elements, too, encode, count and speak, he is not seeking to extend the human meanings of these terms to the non-human world. That would indeed–as was rightly pointed out by the panel–be anthropomorphic and anthropocentric.
He is doing something more radical and much less anthropocentric: contesting that there is a literal sense of language that is to be identified with human language in the first place. Rather, for Serres human language is one local and distinctive instance of a phenomenon that is in no way properly human, but that belongs to all entities. Crystals do not speak with syntactic human language (whoever suggested they did? After all, humans do not undergo an existential crisis when we realise that we don’t transmit information in the same way that crystals do, so why would we assume the reverse?), but it is still quite true and not in the least metaphorical for Serres to insist that crystals speak and human beings speak. (In this he is very close indeed to Jean-Luc Nancy’s notion of ‘renvois de sens’ between all entities in the world in L’Adoration).
Serres’ rethinking of language is intended to be a challenge to the privileging of the mode of syntactic language as the paradigm of which all other instance of language are metaphors, not a universalisation of this paradigm. In this sense, it is akin to a Wittgenstinian argument about definitions: when we say ‘language’ we must take it to mean any emission, storage, processing or reception of information, not only what humans do in communicating with words.
Where does this leave human responsibility and human distinctiveness? Smoothed away on some universal ironing board of flat ontology? Not a bit of it. Serres can (and does) retain a notion of human distinctiveness—he is a self-avowed humanist, after all—but it is a distinctiveness traceable to a quantitative difference, not an ontological difference between the human and, say, the crystal.
There are different sorts of difference, and we must be careful not to fall into the sloppy binary of ‘no difference at all’/’irreducible ontological difference’. Serres’ account is more fine-grained than that. To reconstruct it quickly (and elaborate it beyond what Serres says himself, to the best of my knowledge of his work):
- Both the human and the crystal receive, store, process and emit information. There is no ontological difference.
- But there is an indisputable quantitative difference: the amount, variety and complexity of information that I receive, store, process and emit, is many orders of magnitude greater than the crystal (though the crystal can also perform some tasks of information storage, processing and emission better than I can).
- This quantitative difference between humans, animals and crystals tips over, in a Hegelian way, into what is to all intents and purposes a qualitative difference: my information processing is so different to a crystal’s that, looked upon in terms of their results, they hardly seem the same thing at all.
- This quantitative-qualitative difference does the heavy lifting that ‘human distinctiveness’ is usually called on to schlep: humans bear a peculiar weight of responsibility, not because they are ontologically or at bottom qualitatively unique, but because they are quantitatively (and therefore qualitatively–think Hegel again) set apart within the entities of the world by virtue of HOW they process information and HOW MUCH they process, not THAT they process it.
- So we can, after all, draw a line between human speech and crystal speech, and we don’t need to see fairies under every rock (though the post-Baconian fact-value distinction also needs interrogating afresh in the light of Serres’ ontology, and the odd ‘fairy’ here and there might not always be unwarranted).
- And the question is not whether there is a line to be drawn, but the nature of that line, and the point at which a difference of quantity tips over into a difference of quality.
 Incidentally, note the inversion of order of ‘we’ and ‘they’ in the third element of this ternary rhythm and its subtle reinforcement of the argument being made: Serres is on his game!
 If Serres’ account is to be challenged, it is at the point of the adequacy of this information paradigm, not at the point of the denial of distinctive human responsibility. Can the reception, storage, processing and emission of information act as a sufficient (let alone necessary) paradigm for human being in the world (or, for that matter, for crystalline existence)?