I am still experimenting with the diagram below, but as I was teaching the concept of translation in Harman’s Prince of Networks today, I found it to be a useful heuristic device for thematizing just what is new or interesting in Latour’s concept of translation. Scroll past the Scribd diagram for a bit of commentary.
Clearly I have adapted this diagram from Hjelmsleves model of the sign. All of us are familiar with the relation between the signifier and the signified in Saussurean linguistics (to the left). In naive theories of linguistic translation (NTTs), the idea is that the concept remains the same (content), while it is only the signifier (expression) that changes. There are any number of reasons that this concept of translation is mistaken. I outlined some of these shortcomings in a previous post, so I won’t repeat them here. Latour’s concept of translation is broader than that of translation as it applies to linguistics or the transposition of texts from one language to another. The key point to take home from his analysis– and he doesn’t spell these implications out himself –is not so much the fact that a translated text always differs from the text that it translates, but rather that the process of translation produces something new, regardless of whether the relation is between texts in different languages, conscious minds to world, or relations between objects. What Latour wishes to do, I think, is generalize the concept of translation, such that translation is no longer restricted to the domain of language, nor requiring the involvement of living beings of some sort, but rather involves any relations among actants, human or nonhuman, living or material.
Hjelmslev’s key innovation in the domain of linguistics and semiotics was to recognize that both the plane of expression (loosely the signifier) and the plane of content (loosely the signified) have a form and substance that can enter into different relations with one another. Here I am partially basing my analysis of Deleuze and Guattari’s treatment of Hjelmslev’s model of expression and content as developed in “The Geology of Morals” in A Thousand Plateaus. This discussion would require a far more developed analysis than I’m capable of giving at the moment. For those who are interested, it would be worthwhile to refer to DeLanda’s early work on this essay (here and a number of Delanda’s articles, podcasts, and talks can be found here), as well as the first chapter of A User’s Guide to Capitalism and Schizophrenia by Brian Massumi. While I don’t entirely share the ontological commitments of either of these thinkers, their works nonetheless provide some pointers in the direction I’m thinking.
read on!
Hopefully I’ll have more time to elaborate on the diagram above in the near future (I’m in a rush now), but to understand what’s at stake it’s helpful to take a brief detour through Aristotle’s four causes. I apologize for the inelegance of my diagrams. Hopefully my diagram of Aristotle’s four causes in the upper right hand corner of this paragraph (click to expand) will convey some of the sense of his sorting. The important point to keep in mind here is that Aristotle’s term “cause” (aition) is closer to what we might mean by “reason” than how we think of “causes” today. Each of the four causes is a way of answering the question of what and why a thing is. The efficient cause is therefore that by which something is produced. The material cause is that out of which something is produced. The formal cause is the structure or pattern of a thing. And the final cause is the goal or that for the sake of which something is produced. It will be noted that bisecting the four causes is a dotted line distinguishing what is potential (δύναμις) from what is actual (ἐνέργεια). If the material cause and the efficient cause are associated with potentiality, then this is because matter– for example clay –has the potential to take on many different forms through the agency of an efficient cause. Likewise, if form and finality are associated with actuality, then this is because structure or pattern (the formal cause) indicate that matter has taken on a determinate form, whereas something is actual when it reaches its goal or telos. I take it that these distinctions are well known, so I will not elaborate on them in greater detail here with the proviso that much more can and should be said.
Closely associated with the distinction between matter and form and potentiality and actuality is the distinction between passivity and activity. In a crude version of the Aristotlean schema matter is the passive principle and form is the active principle. Returning to the Saussurean schema of the sign under NTT (the naive theory of translation), the signified as the constancy of meaning is the form or active principle that in-forms the signifier or passive material of expression. This reveals another dimension of form: it is what remains constant or identical in all of its instantiations in matter. The role of matter is simply to take on form, without contributing anything to form beyond the mere instantiation of that form in an existent. Referring to another inelegant diagram in the left above (click to enlarge), we can call this the ontology of “sovereignity”. The sovereign can be anything from a king to a general to a father to God to a boss to a teacher. The sovereign functions as the efficient cause containing the form as an ideational structure (like an architect’s or engineer’s blueprint) that is then imposed on a passive matter. The key point here is that causation is conceived in a unidirectional fashion, passing from the sovereign and his blueprints to the passive matters to be in-formed.
It is now possible to discern the innovation in the adapted Hjelmslevian schema. In the relation between Actant1 and Actant2 there are arrows between the two schemas. These arrows indicate one actant acting upon another actant. It will be noted that both actants possess both a form (structure, pattern) and a substance (a “materiality” broadly construed). As Harman often puts it “there is no such thing as un-form-atted matter. In addition to each actant possessing a form and a matter, each matter contains both a content and an expression. Here content– and I need to say much more about this –can be understood as the other actants that an actant has “appropriated” to constitute itself as an actant, while “expression” can be understood as the manner in which the actant has “actualized” itself qualitatively at a particular point in time. In Harman’s language, content can be understood as the “withdrawn” being of an actant, while expression could be understood as the “sensuous vicar” by which this withdrawn being is expressed for another being. Why, then, the additional dimension of matter for each of these actants? Because in addition to the internal composition of each actant or its content (what I call the endo-consistency of a being which is roughly analogous to Suarez’s “substantial forms”), the being of any actant is infinitely decomposable into other actants or entities. With great caution we can refer to this “matter” as “hyper-chaos”, so long as we note that this hyper-chaos is structured and that its apparent “disorder” is only disorder from the standpoint of the structured being of a particular actant (more on this another time).
The diagram at the beginning of this post requires a great deal more commentary than I can give it here. The shift from the Aristotlean model where we have– at least in thought –pure unformed matter that is completely passive or a sort of “hyletic flux”, thereby requiring in-forming-ing from another actant to my adapted version of Hjelmslev’s model where all actants have both a form (structure, pattern) and a substance (formed matter) initially appears slight. However, if we refer back to Aristotle’s model the significance of this slight shift becomes apparent. If each actant involved in an inter-act-ion has form, then it follows that the actant receiving the action of another actant cannot merely be a passive matter taking on the form of the other actant (the ontology of sovereignity). Rather, because actant2 itself has form, structure, or pattern, it too is an active principle. Yet as an active principle it too must contribute difference. Yet, if this is the case, then it follows that the form of actant3– the outcome produced by the interaction of actant1 and actant2 –cannot merely be the instantiation of the form of actant1, but must instead be a “synthesis” of the forms of actant1 and actant2 producing something new through this inter-act-ion. A task and a critique are here announced at the ontological level. The critique would be a critique of all those vestiges of the ontology of sovereignity where some set of actants is treated as consisting merely of passive materials that take on the form of some other actant. The positive task would be to trace these imbrications of forms in inter-act-ion, investigating the manner in which they produce new forms as a result of the “struggle” between these forms. It now becomes clearly why the alternative ontology is a horizontal, flat, immanent, or networked. No longer can one actant stand apart from the rest imposing a unidirectional, form-bestowing causality on all the others. Rather, the so-called “sovereign” now becomes an actor in a field of actors where causality is bi-directionality and where form is a result of inter-act-ions among actants rather than an identity preserved across chains of inter-act-ions like a signified behind a signified that is in-different to its instantiations in other matters.
November 24, 2009 at 6:27 am
[…] any case, HERE IS LEVI’S LATEST POST, which touches on enough interesting ontology to occupy the rest of your day if you’re in the […]
November 24, 2009 at 8:14 am
Just a quick question. Does all translation involve an actant1 acting upon an actant2 to produce an actant3?
If this is the case, then it would seem like every interaction produces a new actant, rather than producing a change in an old actant. This would seem to be closer to the actualist story that Latour puts forward and Graham criticises, wherein any change to an actor produces a brand new actor (with a brand new form). However, it seems like this isn’t the position you want to occupy (you want some variant of a potentialist position). You seem to want things to in some sense remain the same between interactions. If so, you need to be able to describe translation in a way that doesn’t presuppose the production of a new form, which as far as I understand implies the production of an entirely new actant.
November 24, 2009 at 4:34 pm
Hey Levi,
Perhaps you could help me out a little here:
1) If translation is a process then isn’t it an action? But if it is an action, then like an act-ant, isn’t it also an object? Then technically your relation between actants above explodes from merely three actants to at least four or more.
In other words, we would have to take into account not only the product, or Actant (3), of this inter-action between Actant(1) and Actant(2), but also the process of translation itself – Actant(4) – and any differences Actant(4) produces/is produced by – leading to an Actant(n).
2) Doesn’t this new definition of form get rid of the object that ONLY makes a difference internally – that is, without any exo-relations? Because if I remember correctly, in an earlier discussion you said that onticology didn’t discount the object far off in space that merely had endo-relations (since technically it was still making a difference). Yet, it seems that form described here then would always require a exo-relation acting upon it.
A preemptive thanks for any clarification.
November 24, 2009 at 5:28 pm
Nate,
I don’t really see why we’re required to treat interactions between actants as itself being an actant. Responding to Pete’s question, it is not necessary that every relation between actants produces a third actant in the sense of a discrete entity. Such a relationship could just produce a modified state within the second actant itself.
As for your second question, I think you’re conflating two distinct issues. Translation is always and only a relation between two or more entities. There is no instance of translation that does not involve two or more entities. If it helps to visualize what is going on here, just think in terms of black boxes: actant1 (input) —-> actant2 (black box) —-> product (output). That’s all there is to it. Think about your phone. You have an input (electrical pulses), a black box (the phone itself), and the product (the sounds that come out of the receiver).
Nothing that I’ve said here about form undermines the independence of actants or that they can act without acting on acting on anything else. The fact that actants can and often do enter into exo-relations in no way undermines the point that actants have an endo-relational structure of their own. More importantly, questions about one entity modifying another entity through entering into exo-relations with it in no way undermines entities possessing internal dynamic structures of their own. If you reread the post I think you’ll note that to have form is to be active. This is precisely why actant2 is not merely a passive form receiving the impress of another actant’s difference (that was, in many respects, the point of the post). I would go so far as to say that transcendentally this endo-relational structure, dynamism, and autonomy is a necessary condition for understanding what takes place in inter-object interactions and why they produce outcomes that are new and not merely rote copies of the imprints or differences they receive.
I would also add that to make a difference is to be a generative mechanism… A point that I’ve had a great deal of difficulty getting across to folks that want to treat difference as either “differing from” (attributing a diactrical structuralist position to me which I don’t advocate) or who conflate producing differences as being equivalent to “being distinguished” (among folks who seem unable to envision generative mechanisms– i.e., one half of the split in split objects –as not being about a consciousness distinguishing two things– usually inaccurately in ontological terms –from one another). At any rate, you wouldn’t have much of a generative mechanism if generating (making differences regardless of whether anyone is there to perceive them or any object there to be acted on by them) weren’t, well, generating!
November 25, 2009 at 2:29 am
Thanks Levi,
Some further thoughts:
http://un-cannyontology.blogspot.com/2009/11/ubersetzung.html
November 25, 2009 at 6:11 pm
[…] to a couple of my posts from earlier this week on translation, Nate over at Un-canny Ontology writes: What is translation? […]
November 26, 2009 at 2:15 am
This is brilliant–a great explanation. Yay Hjelmslev! And thanks!
May 23, 2010 at 12:12 am
[…] have called, following Latour, this process by which one object grasps another object translation. In this context, translation does not refer to a phenomenon unique to language, though the […]