Week 2: Cognition in High Frequency Trading
Audio introduction to this weeks discussion here.
Last week, we began our reading group on The Digital Condition with a text that made no reference to digital technology. Perhaps it was inevitable that in the ensuing discussions the question of (and call for) examples kept coming up. This weeks reading, Cognition in High-Frequency Trading: The Costs of Consciousness and the Limits of Automation by Armin Beverungen and Ann-Christina Lange, explicitly presents itself as dealing with High-Frequency Trading (HFT) as an »exemplary case for the study of algorithmic cultures« (pg 2, also here). There is a lot of exciting literature on what exactly an example is and does out there, but this is not the place to go down this particular rabbit hole. Suffice it to say what everyone engaging with theory already knows: examples (as distinct from applications) are always trouble and vehemently refuse to play the illustrative role assigned to them in Christian Wolff’s fable of enlightened rationality.
In Beverungen and Lange’s paper, HFT functions as an example not once, but twice: HFT is an example of algorithmic culture and we can investigate questions of affect, cognition and automation accordingly. But HFT serves also as an example of an example, precisely because HFT is already a well-established »illustrative example« (5) and presented as »emblematic« (2) of these issues. They write:
»As media and cultural studies more broadly engage with how the computational media that surround us constitute a ‘technological unconscious’, and how algorithms condition our lives and cultures, HFT is often presented as emblematic both of fully automated computational systems and of affective contagious dynamics that mark these systems. Key here is the assumption that algorithmic HFT, working at speeds below the threshold of human perception, escapes the cognitive capacities of human traders, so that affective dynamics mark both the human-machine but also the machine-machine relations, as in the case of the flash crash. In this paper, we want to contribute to this literature with a more nuanced account of affective contagion, cognition and automation in HFT.« (2–3)
So HFT is not only an incredibly exciting case study by itself, it also reveals a lot about contemporary media and cultural theory, the way it deals with ›examples,‹ and its assumptions regarding the technological unconscious. It is particularly for this reason that I have decided to add this paper to the syllabus for an interdisciplinary reading group, because it introduces a whole range of current debates while simultaneously pointing to crucial shortcomings. For those not familiar with the field, I want to stress that Beverungen and Lange don’t construct weird strawmen or dig out obscure takes on our contemporary condition, but genuinely stick to approaches well-known in media and cultural theory. In this short introduction to this week’s reading I want to highlight three brief points where the example of HFT helps us to engage in more careful theorizing. In doing so, I am effectively only highlighting half of the paper. While it contributes to both, »the social studies of finance through a close engagement with media and cultural theory concerned with the technological unconscious and cognition, and to media and cultural theory through a more thorough empirical encounter with the phenomenon of HFT« (3), I am predominantly stressing the latter aspect for this reading group.
Yet, as the first aspect is nevertheless of particularly interest to my own thinking, I want to briefly remark on the relevance of the former point. Recently, especially since the financial crisis of 2007–2008, there has been a spike in interest in the »dialogue between the research of media and the research of capitalism« which, at least in the German context, is arguably rather unusual. In various places the question is raised what media theory — as media theory — can bring to the table in such a dialogue. In addition to calling for a more nuanced account of affective contagion, cognition and automation, the paper offers a quite convincing interface for facilitating such a dialogue. Particularly thinking about the automatisms of capital from a media theoretical point of view might be an incredibly productive approach, both historically and critically. With this in mind, let me now return to the empirical encounter with HFT and its implication for media and cultural theory.
Hammers and Algorithms: Marginal does not mean absent
Algorithms in HFT are clearly not like a hammer that you pick up because it would be quite painful to hit the nail with your palm. Granted, in this new scene playing out on the trading floor the human no longer seems to be the sole protagonist, carrier of all meaning just as she carries the hammer. Yet, while it might seem »fair to say that media technologically-armed financial markets constitute machine-machine ecologies in which human traders play an increasingly marginal role« (2), the paper vividly highlights how »central humans remain even to HFT« (16). In fact, the paper is structured around a variety of strategies and processes by which humans still remain in the picture — and not just accidentally — even within »the complex ecology of the financial markets« (12). This is not only due to the various strategies of establishing cognition (8–12) and the limits of automation (12–17), but also because HF traders are »actively partaking in the construction of the computational environments and cognitive systems in which they engage« (3).
While the image of the completely passive consumer (the one known mainly for helplessly watching TV programmes) has largely been dispelled from the media-theoretical debates, the image of the user interacting with this strange and increasingly complex technology-thing has taken its place as the epitome of the human figure in a technological ensemble (and perhaps gives the image of the consumer a second, rebranded life). But, as they show, when dealing with HF traders, the idea of ›consumers‹/›users‹ proves wholly insufficient.
Beyond questions of the relative centrality/marginality of human participation, to me the paper also — on a more general level — points to other ways of conceiving participation in an event beyond the swinging of a hammer. As one of the interviewed traders remarks: »It’s an incredible amount of responsibility I bear« (15). The modes of participation might change, but that does not necessarily imply a complete absence.
Ontology of time or labour of consciousness?
A lot of the recent obsession with questions of microtemporality comes from the fact that — as they point out with reference to N. Katherine Hayles — »consciousness is slow relative to perception« (5). It’s predominantly an argument about speed, particularly about speed relative to human capabilities. Yet, what begins as an argument about speed with reference to the human is often dropped in consequent arguments and we end up with some fundamentally other, non-human temporality, beyond (or rather, below) all access to human cognition. It gets theorized in terms of affect, contagion and the inhuman. With HFT, however, we can think about this more carefully, without denying that there still is something worth considering here:
»humans are too slow in acting in markets, and their cognitive capacities by themselves are too limited for comprehending financial markets. … Certainly HFT takes place in milliseconds, and therefore in the cognitive gap. Yet this does not necessarily mean that HFT takes place outside of human control and comprehension. As we will outline below, our empirical research suggests that humans precisely devise algorithms as cognizers supporting their cognitive capacities for comprehending the market« (8–9)
Thinking about various modes of awareness and — crucially — the labour that goes into establishing forms of comprehension, cognition, and control, might be a useful approach that gets away from an ontologizing discourse about different temporalities (see esp. 10).
An Exemplary Case?
In this short introduction to the week’s reading, I have focussed on HFT’s status as an example of an example for techno-ecological thinking, and how Beverungen and Lange’s paper complicates this. To end, I finally want to return to the question of HFT as an exemplary case. Last week, we briefly talked about the scenes from which a theory is generalized. Is our thinking on technology and sense shaped by a scene of a human holding a hammer and doing things with it? Is it based on the experience of making a compromise with the materials when weaving a basket, the complex transfers of work/energy when moulding a brick, the production and distribution of the free software movement or the adaptive interaction when riding a horse? What is the exemplary case for a techno-ecology? Maybe HFT is a contender. It sure fits all the hallmarks of techno-ecological thinking. But it is so clearly a political, economic and institutional scene, inextricably part of the history of the »automatisms of capital« (17) and the »financial markets as cognitive systems« (3), that it might spare us from the eternal recurrence familiar from any seminar and reading group: Somebody makes an argument ›about technology,‹ somebody interjects that ›Yes, but also economics/politics/institutions!‹, everyone — rightly! — nods and agrees, because, of course also those things, and we move on to the next argument about technology. If we keep HFT as an exemplary scene in the back of our mind, maybe the question of technology presents itself differently.
