Sunday, February 9, 2014

What's the case for sousveillance? (Part One)

Steve Mann - Sousveillance Advocate

Steve Mann (pictured) has been described as the world’s first cyborg, and as a pioneer in wearable computing. He is certainly the latter. I’m not so sure about the former (I believe Mann rejects the title himself). He is also one of the foremost advocates for sousveillance in the contemporary era. Sousveillance is the inverse of surveillance. Instead of recording equipment solely being used by those in authority to record data about the rest of us, sousveillance advocates argue for a world in which ordinary citizens can turn the recording equipment back onto the authorities (and one another). This is thought to be beneficial in numerous ways.

I’m interested in whether there is a strong case to be made for sousveillance, particularly in light of the increasingly prominent role of data-monitoring in our lives. Fortunately for me, Mann has recently released two papers that develop the case in several ways. One of the papers sets out a series of economic and social justice arguments for sousveillance; the other develops a framework for thinking about different types of “veillance” in society. I want to analyse both papers in this series of posts.

I do so with some trepidation and with a forewarning to the reader. I wouldn’t usually say this — I prefer to see the good in everything — but on this occasion I fear I must: Mann’s papers are not of the highest quality. They are strangely written and poorly focused, sometimes engaging in opaque and incomplete argumentation, and sometimes going off on strange etymological and historical tangents. I’m going to try my best to focus on the main arguments, and to reconstruct them in as charitable a way as I can. Nevertheless, I will occasionally point out some lacunae and weaknesses.

With that warning out of the way, I shall proceed. This post (and the next) will deal with Mann’s economic/social justice arguments for sousveillance. It starts off with some definitions and clarifications. It then looks at Mann’s claim that sousveillance can facilitate beneficial economic exchanges.

1. Definitions and Clarifications: The Wrong of Surveillance?
Mann seems to love etymology. Both of his papers are riddled with odd excursions into the etymology of particular words (surveillance, sousveillance, terrorism, economics and so on). There is no clear sense of why this is done. I find etymology pretty interesting myself, but I’m inclined to think it is somewhat of a distraction in this instance: I don’t think it gives us any real insight into the phenomena in question. Nevertheless, if we are going to consider the case for sousveillance, and if sousveillance is introduced by way of contrast with surveillance, we need some definitions in place at the outset (even if they are purely stipulative in nature).

Fortunately, Mann (and his co-author Mir Adnan Ali) oblige. They define veillance as the watching (or recording) of a person. This can be done through video cameras, but the definition of “watching” is not restricted to the visual sphere. The collection of any personal data that can be recorded and transmitted will count. From this core concept of veillance, the definitions of sur- and sous-veillance arise. As follows (from p. 243):

Surveillance: Monitoring undertaken by an entity in a position of authority, with respect to the intended subject of the veillance, that is transmitted, recorded or creates an artifact (i.e. like a digital recording or video).
Sousveillance: Monitoring undertaken by an entity not in a position of authority, with respect to the subject of the veillance, that is transmitted, recorded or creates an artifact.

Mann and Ali are clear that “authority” here is understood in terms of ability and legitimacy. In other words, a person possesses authority over another if they have the ability and the legitimacy to impose their will on that other.

This is the first slip-up in the argument for me. They explicitly say that legitimacy is understood in a “normative sense”, but I don’t see why they say that. Indeed, many of the most problematic cases of surveillance — ones that sousveillance may be able to counteract — arise precisely because the person doing watching can illegitimately impose their will on another. Furthermore, depending on how you define legitimacy, this definition of “authority” risks foreclosing much of the ethical debate about surveillance and sousveillance. If legitimacy entails the moral right to enforce your will, it’s difficult to see how or why the use of surveillance equipment would be of major ethical concern. I would suggest, then, that we drop “legitimacy” from the definition of authority.

This raises the next issue. It seems obvious that in making a case for sousveillance you must, implicitly or explicitly, believe that there is something morally problematic or sub-optimal about a world in which surveillance dominates. But what is that something? Well, first of all, let’s consider the advantages of surveillance. Clearly, surveillance has advantages from the perspective of authority. It can be used to police and enforce behavioural norms (e.g. street cameras and laws against vandalism) or to prevent the breach of such norms. Consequently, to the extent that these norms are morally valid, surveillance is of benefit to us all. The obvious disadvantages of surveillance are when it goes too far, and personal rights such as the right to privacy are traded-off against the good of enforcement, or when the norms being enforced are not morally valid.

Another, perhaps more subtle, problem with a surveillance culture is best-expressed using republican conceptions of liberty and non-domination. One thing that constant surveillance seems to carry with it is the implicit threat that if you do something to displease the de facto authority figure, you risk punishment or sanction. You live in the permanent shadow of a threat. This will force you to engage in ingratiating, self-censoring and extra-cautious acts. The position strikes me as being similar to that of the happy slave in neo-republican political theory. The happy slave is happy only to the extent that he or she doesn’t step out of line. That’s not real freedom (according to the republican theory). Those us living under the domination of surveilling authorities might have a similarly restricted type of freedom.

This problem of surveillance and domination needs further exploration, but it seems important to me. Indeed, I think it can be used to great effect when evaluating the strengths and weaknesses of Mann’s case for sousveillance. Let’s turn to that case next.

2. Trust, Exchange and the Case for Sousveillance
The main argument that Ali and Mann make for sousveillance is based on the value of efficient economic exchange. As any classical economist will tell you, free and fully-informed exchanges between rational agents should increase societal well-being. The idea being that such a system of exchange ensures that resources are distributed to their highest expected value uses. Now, there are many problems with this model, particularly in terms of the idealistic assumptions one needs to make in order for the conclusion to hold. Nevertheless, Mann and Ali’s arguments are based on the notion that sousveillance gets us closer to those idealistic assumptions.

Central to this argument is an analysis of the conditions for efficient economic exchange. It has long been clear that social cooperation can be mutually advantageous. It has also long been clear that such cooperation carries risks. Hume’s story of the two corn farmers illustrates the point rather nicely. Imagine that there are two farmers, A and B, both with crops of corn. These crops will ripen at different times. Each farmer will need the help of the other to ensure that they can harvest their crops, and put them in storage in good time. Without such help, a portion of the crops will start rotting in the field. In this case, cooperation would be mutually beneficial. The problem is that a purely rationalistic analysis suggests that they won’t help each other: if farmer A helps B before his crops ripen, then farmer B will have no real incentive to help farmer A later on. Reasoning backwards, A will expect B to betray him and so won’t bother helping B. This is sometimes referred to as the Farmers’ Dilemma.

There are various solutions to this dilemma. A legal system that enforces promises is one: if the breach of promise carries with it a risk of legal sanction, then more people might be inclined to keep their promises. So too is trust: simply voluntarily committing yourself to help another, in spite of the risk. Some people argue that trust is a social emotion that evolved so that we could solve the problem of social cooperation. And some people argue that trust of this sort is incredibly virtuous. Something that society should be keen to promote and protect. Indeed, trust plays a considerable role in many important social exchanges. People can feel offended if you don’t trust them, and may back out of an exchange if you seem to lack trust. Think, for example, of how betrayed you might feel if you caught your partner snooping through your text messages just to make sure you were being faithful.

Ali and Mann argue that sousveillance can facilitate beneficial social exchanges. They say it does so by making the parties to an exchange less vulnerable to exploitation. The mechanism for this is not stated, but I assume the authors are imagining something like the following: There is a system in place that will enforce promises if they are breached (this system could be a formal legal system or an informal social one). This system relies on proof of claim before enforcement. Those who use sousveillance will record every detail of every negotiated promise. In this way, they will always be able to prove a claim should they need to rely on the social system of enforcement. Consequently, every exchange with a sousveiller will carry an implicit threat of enforcement. If both parties are sousveillers (as they should be according to Ali and Mann), they can keep each other honest, and thereby clear the path to beneficial exchange. In this sense, sousveillance is a trust-substitute: it overrides the vulnerabilities inherent in exchange without forcing us to voluntarily assume the risk of exploitation.

To lay this out more formally:

  • (1) If mutually advantageous social exchanges carried less risk of exploitation, people will be more likely to undertake them.
  • (2) Sousveillance helps to reduce the risk of exploitation inherent in mutually advantageous social exchanges.
  • (3) Therefore, sousveillance increases the likelihood of people undertaking mutually advantageous social exchanges.

We’ve already considered the case for premise (2) in the preceding paragraph. I want to dwell on premise (1). It seems to me that this is potentially vulnerable to a counterargument. The counterargument brings us back to the virtue of trust. Although some exchanges could be facilitated by sousveillance, it could also be the case that people insist on the voluntary assumption of risk as a gesture of good faith prior to entering into an exchange. We could imagine, for example, one of the CEOs of two major corporations, negotiating a merger deal, asking the other to “switch off” his sousveillance equipment before they agree on the final terms of the deal. After all, if this merger is going to work, they need to trust each other, and they can’t do that if they are constantly monitoring and recording one another’s words.

This might, of course, be terribly naive, and agreeing to this gesture of good faith may be costly, but humans are sometimes irrational and one could imagine this kind of insistence taking place. What matters is whether the number of valuable exchanges in which people insist upon trust, will be larger than the number of valuable exchanges facilitated by sousveillance. In their discussion, Ali and Mann break the possible exchanges down into three categories: (i) those that are unaffected by the presence of sousveillance; (ii) those that are facilitated by sousveillance; and (iii) those that are discouraged or prevented by sousveillance. As long as categories (i) and (ii) are larger and more valuable than category (iii), a case for sousveillance can be made.

So which is it? Ali and Mann try to argue that the number of exchanges in category (iii) will be minimal. They do so on the grounds of desensitisation. As sousveillance becomes more widespread, people will adjust their expectations to accommodate it. They will be less “creeped out” or offended by its use. Arguably, this has already happened with surveillance technologies. Whenever I get the train, I see little signs reminding me that my every movement is being recorded by CCTV. It doesn’t bother me. I’ve become so used to it. Why wouldn’t the same thing happen with sousveillance?

I think there are some problems with this argument. While it is true that we could become desensitised to sousveillance if it achieved sufficient social penetration, that seems to assume what needs to be proved, namely: that sousveillance will achieve sufficient social penetration. Elsewhere in their article, Ali and Mann defend this on the grounds of economic inevitability: sousveillance will be so economically beneficial that it will become widespread. But, again, that seems to assume what needs to be proved: that sousveillance really is economically beneficial. If the economic benefit of technology depends on whether it facilitates voluntary exchange between parties, and if a sufficient number of parties are offended by the use of sousveillance technologies, then they won’t be economically beneficial. People won’t consent to their use. This is markedly different from the surveillance case. Since surveillance technologies are imposed from the top-down — by de facto authorities — they don’t require the immediate consent of those being watched. Economic exchanges arguably do.

Admittedly, this is a technical objection, based more on how Ali and Mann make their argument, than on what I think the reality is going to be. The fact is that the marketplace is currently characterised by inequalities of bargaining power between parties to economic exchange. It is perfectly possible that those inequalities create conditions in which veillance technologies will get a foothold. This may clear the path to widespread sousveillance

Okay, that’s it for part one. In part two, I’ll consider the remainder of Ali and Mann’s economic arguments for sousveillance. I’ll then turn to their social justice argument, which is based on an analysis of the role of bureaucratic power in modern society.

No comments:

Post a Comment