publications
In the last instalment of this series, I introduced the notion of "sociotechnology". In the current paper I will discuss this notion further. My goal is relatively modest - to simply sketch out the concept and its scope and limits. Applications of the idea will come in future work. I divide the following discussion up into five parts:
  1. Recap of what sociotechnology is
  2. Relationship between sociotechnology and science and between sociotechnology and crafts
  3. Ethics and technology in general
  4. Ethics and technology in specific - the case of sociotechnology
  5. Why this matters

What is sociotechnology?


Loosely speaking, a sociotechnology is a field of study that makes use of scientific findings about social behaviour in order to modify this behaviour or prevent such. (From the scientific perspective these are the same sorts of problems. Those of the readers who may remember Newtonian mechanics know that one can accelerate a body to a greater velocity by applying a force to it, but also know that we can stop or slow down a body the same way by applying a force in the opposite direction.)

Examples that exist currently are some parts of so-called management science, criminology, action theory, and normative economics. The latter deserves a special comment: I shall return to the question of what economics is about in the next section.

There are many disciplines that are protosociotechologies. This weird word means they are fields of endeavour and research that ought to (in my view) make use of scientific findings to inform action, but are underdeveloped in this respect. The law is the biggest one. Forensic psychologists tell us that our legal systems do not adequately reflect what is known about the unreliability of eyewitnesses, how the so-called "lie detector" is nothing of the kind, and many other interesting findings. Another protosociotechnology is normative international relations, diplomacy, and foreign policy. It is a truism that we should study how power arrangements work in a country before planning any sort of intervention there, whether be opening trade agreements or invading them. That this is not done adequately, and that ideology trumps research is why this field is still underdeveloped. Of course, political science (properly understood) is itself in many respects underdeveloped so it is not surprising that the corresponding technology does not amount to much as yet.

Enough examples. Let us examine sociotechnology in more detail.

Relationship between sociotechnology and science


As stated above, science is used by sociotechnologists to form plans or artefacts. In this case, many of the artefacts are social systems or pieces thereof rather than machines, drugs, or buildings, but they are artefacts all the same. (Mario Bunge attributes a saying to Ravel: "I am artificial by nature" - a description of humans that seems to work well enough.)

Sociotechnology differs from what one might call social crafts. A craft, on this understanding, is a nonscientific artefact. For example, canoes traditionally built by the natives of North America are wonderfully made artefacts, but they were not made informed by scientific knowledge. (A characterization of what science is found in the first of these papers.) By contrast, a canoe made with an understanding of the strengths of materials garnered in materials science, or made to minimize the devastation of birch trees that was obtained by studying their growth cycles is a product of technology.

It is of course not sufficient (but by definition it is necessary) that science be used to form a sociotechnology. Let us take a look at why this is so. Consider any finding of scientific research. Last time I used repeatedly the example of the linear relationship between unemployment and criminality. We saw that this could be written as an equation of the form C=a+bU. Now what? What intervention is desirable, assuming this relationship to be approximately true? Many of us may suggest that criminality be lowered by lowering unemployment. But notice that the relationship we have found does not tell us anything about which way to manipulate. (Strictly, one should have a few more well established relationships on hand before intervening at all, but let us keep the simple case for the time being.) If for some reason we wanted to increase criminality - for example, maybe we own one of those private prisons in the US - we could try to work for an increased unemployment. This moral neutrality - that a law statement can support two distinct (at least - sometimes more, depending on the degrees of freedom of the relationship) proposals for manipulation - is universal to the findings of scientific research. The moral neutrality of science is its greatest strength, as it can be used to skewer the illusions of anyone, regardless of her viewpoints. Or it can be used to simply describe and explain the way the world is.

Richard writes concerning moral neutrality: Then comes a reasoning for the "moral neutrality" of forms of expression of the type C=a+bU. "This moral neutrality [exists due to the fact]that a law statement can support (at least) two distinct proposals for manipulation". One has the choice either to decrease or to increase the crime rate. Both opposite choices can be supported by the same law; each has equal scientific weight. Hence neutrality between choices. Let's get the thing into focus. The equation is the written expression of a significantly stable relationship between two sets of social facts transcribed as crime rates and unemployment rates. It is a bit of knowledge that tells us something about the world as it is. So far so good. But is it this fact, the fact that the hypothesis is valid, that is true, that makes the hypothesis morally neutral. No. What makes it morally neutral is that anyone who uses the hypothesis as a reason, as the grounds, to do something has a reason of the same weight as someone who uses the hypothesis for the opposite action.> Now suppose the hypothesis is junk. Its importance to she who uses it as evidence to back her view is no more or less than its importance to a rival who uses the same evidence in support of the opposing view. If it's good support for one it's equally good for the other; if it's bad support for one it's equally bad for the other. There's indistinguishable moral neutrality in both cases. Following the observation that the neutrality is derived from the sameness and not the equality of the supporting of the supporting evidence, note that flipping a coin to decide between two related opposite choices has the same moral neutrality outcome as have your law statements. It just doesn't sound as impressive. Let's probe a bit further. Someone comes up with C=c+dR where C is crime rates and R is Blacks. Let's also say this hypothesis has passsed the law statement exam. Lets add that the new hypothesis is as relaiable as the first, equivalent to it as far as science goes. With both equally strong hypotheses in hand, she, a private prison owner, wants to raise the crime rate and argues for action aimed at unemployment policy while he wants to reduce the crime rate and argues for amendments to rational laws. By construction each has equal scientific support. The example fulfills exactly the criteria of moral neutrality. We are logically forced to conclude, in this particular case at any rate, that unemployment policies and racial policies are morally indifferent. Of course we all know they're not. But that's forgetting that it's not what it is but how much it weighs as evidence that matters. Whatever the differnt impacts on the lives of different people following from the alternative policies, from the science-approach perspective one's as good as another. The paragraph that gives us this moral neutrality stuff continues with a sentence that begins, "The moral neutrality of science is its greatest strength...". We have already implicitly concluded that moral neutrality for hypotheses in the form C=a+bU has no strength. Moral neutrality as defined there is the tautology that those who use the same reason to support an argument use the same reason. The thing in issue is not a reason in support of a hypothesis but the use of a hypothesis as a reason. The same as we can use 'heads or tails' as a reason. Is the hypothesis strong? If it's the only reason given by everyone that reason doesn't tell us. Nor does it matter. But here a different can of worms is opened, a moral neutrality that has strength, that is strength. This assertion needs an explanation. We get none. Also the subject is no longer law statements but science. How are we to understand the transition from one to another. We can suppose by piling up in one column all law statements of the given kind and in a second their corresponding moral neutralities, then totalling up each column and calling the first total 'science' and the second 'grand moral neutrality'. This is sheer nonsense. But if we keep our minds clear and simply follow the path laid out for us, all we need to say is that when 'science' is used by one to support his position and the same 'science' is used by the other to support her related opposite position, then 'science' is morally neutral. As we've seen with hypothesis, when all use the same 'Science' as reason (grounds), it need not be good science. 'Science' as a reason is no better than 'heads' as reason when playing according to the moral neutrality rule. One can say, as meaningfully as you have put it elsewhere, Moral neutrality is the greatest strength of coin tossing. In the moral neutrality game , the form and content of science don't matter. Science taken as knowledge has no significance, no importance here. We can substitute 'science fiction' without the slightest loss. However, you've changed the game in mid-sentence. [...]"
Keith responds to Richard's remarks about moral neutrality ... As for the discussion of the moral neutrality, this is of crucial importance; so let me start by illustrating another possibility. It might have been that there is no relation at all between unemployment and criminality. Many people have suspected there is; the well confirmed hypothesis I have discussed not only tells them that they (and not their critics) were correct, but a relationship between the two macrosocial variables. If there really had been no relationship between unemployment and criminality, any attempt to manipulate criminality (either way!) by manipulating unemployment would have been unwarranted; it would have been acting on false information. Now, we all act on hypotheses that are to some or other degree false - this is unavoidable. What is avoidable at least sometimes is some error, of some kind. A "meliorist" approach is what I am defending, not an infalibilist one. In this sense asking about the moral neutrality (or not) of a hypothesis known to be very false (i.e. disconfirmed) is inappropriate. It is the approach that is morally neutral and the findings. Since a disconfirmation is not a finding (per se; it is of course a finding of the uninteresting negative result: "Criminality and unemployment are NOT related ..." which cannot support any policy of any kind. Negative findings are in this sense "cheap.") As far as I can tell Richard's criticisms here miss the mark. Hence his coin-flipping example is right as far as it goes only. It is quite correct (in some sense - arguably it is immoral to act on hypotheses of negative warrant, but that I will discuss later) that to flip a coin to decide on a policy is just as morally neutral and doesn't sound impressive. It is impressive, at least to me, however, that we have discovered the relationship between two important quantities of concern to many of us. It is also then impressive that we can make our policy decisions accordingly. If we don't study the world (including the social world) we are essentially making the coin flips that Richard rightly derides when we make any policy decision. Now, how do we decide what to do? Well, then our values come in. This is why I discussed sociotechnology: a way of making our values, including our social ones, explicit. Social scientific research can tell us more exactly our values will have in terms of impact on others. We can then weigh them, deliberate about them, and so on, both with others and ourselves. Saying one doesn't care about flooded peasants in Banania is a moral stance (albeit to me a callous one); we can respond to this stance by calculating and measuring how many lives will be lost or ruined, how many hectares of land will be wasted, etc. Some might say that even one square meter of family farm ruined or life lost is significant and we don't need to measure or calculate to determine that. I agree, but since our lives invariably have at least some impact on others, it stands to reason to figure out as best we can what that is, and then decide whether we can make appropriate trade offs. (Suppose we flood one square meter of land in order to let 1000 people produce better rugs so that the local huts can be less cold during the winters in Banania? Is the trade off so bad now? But wait, the rugs require harvesting a local tree for fibres, so ... and so on.)
Keith expands further on law statements... I thought it was clear in context that law statements are the (temporary, fallible, etc.) findings of scientific research. All of them are equally morally neutral. Science (as we discuss later) does occasionally get corrupted. This is to be expected. However, throwing it out would amount to throwing out all human institutions - we are fallible creatures. Instead I propose being vigilant: scientific misconduct is not very common and the media tends to report on it in disproportionate amounts. Hence we should be able to keep track of it. (Were it to become so common that we couldn't, I dare say our societies would be in such a sorry state that studying them might prove hopeless anyway.) Stephen Cole, a sociologist of science, has studied this matter and scientific research is fairly "clean." One area where it is recognized by some that it is not is in economics, as I discussed later on. And so "science as a reason is no better than heads when playing according to the moral neutrality rule ..." is just beside the point. The epistemic and the ethical are separate considerations.
Richard writes concerning skewering illusions: ... For the whole sentence is, "The moral neutrality of science is its greatest strength, as it can be used to skewer the illusions of anyone, regardless of her viewpoints." We have already seen this can't be entirely true. For if she invokes the same 'science' to support her viewpoints as you use to back yours, then she hasn't been skewered. And if she uses other grounds? Other grounds can only be 'not science'. Let's have in mind scientific knowledge and call everything else ordinary knowledge. You use scientific knowledge as a reason and she returns with ordinary knowledge. It does not logically follow that the first can be used to skewer the second. You say it is not the 'knowledge' in science that is its "strength" but rather "the moral neutrality of science". And that, at the risk of endless repetition, is defined as the episode where one and the same lump of evidence supports two distinct courses of action. Granting this, it may be the case that ordinary knowledge, while in form and content different from scientific knowledge, has the same weight as the latter when used as evidence or grounds. We saw this possibility above when we compared U and R as variables in criminal rate statements. One hypothesis was as good as the other. In this event your skewing optimism is premature. All the more so if ordinary knowledge turns out to be more robust than scientific knowledge when used as reason. Then the skewer would be in the other's foot. At all events we must move the argument step by step without skipping. We can compare the worth of different hypotheses as truth statements through established methods. There are no established methods for comparing the respective worth of scientific and ordinary knowledge (if they can even be got clear). That's beyond our task here and certainly well beyond my means. We do know that science alone can't decide between science and non-science; it's one of the parties in the dispute. And we can kick the can just for fun. Can science skewer all illusions of non-scientific pedigree? I don't know how to formulate the quetsion in a way that it can be answered. But we need not remain silent to avoid the risk of illusion. Or of skewers.
Keith responds to the criticism of Richard concerning illusions and skewers... Richard claims that the form and content of science don't matter, and that we can put in "science fiction" instead of "science". Maybe so from the perspective of form, not so from the perspective of content. It is precisely because we know something about the world through science that we now have (perhaps) our ideas about it skewered. It is true that we occaisonally make mistakes in scientific findings, and once in a while we find that the prescientific understanding of something was the correct one, but until that happens in a specific case, it is the ordinary understanding of something that gets refuted by scientific research. This is also so in every day life; people (ought to!) change their minds when new data are available.

This is why there are two main branches to economics. Above I mentioned that normative economics is a sociotechnology. Normative economics is to be contrasted with positive economics, a social science like sociology, history, political science, culturology, and others. Positive economics is just the study of existing (or historical) means of exchange. (One need not limit economics to a discussion of societies that have used fiat money. Barter, reciprocity agreements, the potlatch, and other arrangements are also economic - at least in part - in character.) By contrast, normative economics is a sociotechnology because it attempts to create economic arrangements of various kinds. (That this is sometimes attempted by leaving things as they are as in the case of right-libertarian economics is beside the point.) Moreover, it is not a criticism of this view to point out that the economics textbooks and popularizations merrily conflate these two notions. The reason for introducing the strange vocabulary of these papers is so that we have increased categorization to better understand the differences between things and processes.

Ethics and technology: a primer

The principle that one can ground rules for action based on law statements found by scientific research is called the "rule based on law principle". The rule based on law principle, as we saw, leads to at least two possible rules based on each law statement. One or more of these might be desirable from the perspective of someone. (Or any other domain of ethical evaluation. It is not my place in this paper to evaluate who should make ethical decisions and what sorts of things: people, other animals, other living things in general, the biosphere, societies, etc. That in itself is a sociotechnological problem, and suffers from a rather annoying bootstrapping issue.)

Because technology is rational action, it necessarily involves plans. (Planning is not however sufficient for technology - a traditional canoe of the sort I mentioned earlier can be sketched or described in a oral tradition that details its construction.) These plans are where the moral commitment is clearly in place. "We propose to do this, with these means, and towards these ends in view." This sketch of what plans contain shows that all three components are morally sensitive - we might select an inappropriate task, means, or ends. Plans in technology have to be detailed but not stifling - not all intended outcomes may occur, nor all outcomes intended. One can of course plan for failure - in fact, one should, morally speaking. For example, in medical research, it is vital that we know what drugs a patient is taking when testing a new one, in order to better avoid synergistic effects and so on. But not all outcomes can ever be foreseen, and so it is also vital that "failure modes" be graceful and minimally damaging as possible. A computer program, which runs out of disk space suddenly, should not erase itself from memory silently and announce nothing to the user.

Richard picks up on the canoe example ... A Native person knows how to build canoes. He does it in traditional ways using traditional tools, techniques and materials. He can refer to no science to explain his work. But it works, and well enough for him. A specialist knows how to build canoes. He uses scientific ways, and so on, and he can refer to the studies that led him to do this and use that rather than something else. The specialist has that S-O Approach. He can teach the other to build an unquestionably better canoe. In spite of all the scientific weight against him, and no scientific evidence for him, the other turns down the offer. That "should" business the scientist came in with didn't take hold. Grumbles and suppressed insults. And yet the Native has an effecitve answer that stands up well to all the scientific grounds raised against him. He simply uses his common sense and says, 'See here. If I do it my way with my means I have full control over what I'm doing. If I do it your way I grant you I get a better canoe but at a cost I am not willing to pay. I become dependent on you for everything to have and to do with canoe building except for labour, and even as labour I have to follow your instructions.' (Especially if the orders come from scientific management, he could have added.) Rewriting this sketch formally we have: M opts for a better canoe and weighs in with scientific data to support his view. N argues for an inferior canoe and grounds his choice on reasons of autonomy. This form is not identical with, but is similar in the most important respects to, your definiton of moral neutrality. There is no scientific way to break the deadlock, for the reason given two paragraphs above. There is nothing in the rules that states every issue has to have, or even is amenable to, a scientific fix. Intuition does not tell us science made the stronger case.
Keith responds to Richard's use of the canoe example ... How to decide when to use scientific findings and when not to is the subject of the fourth part of what I wrote, which Richard has (apparently) not read yet. That's fine, just that much is addressed there. And how to decide between science and not is a matter of philosophy; just let me say here that a non-scientific worldview is possible, though increasingly not so easy. One can remain apart from science, at the "cost" of not having any means to fully evaluate its findings or policies that decide to make use of it - or not! The educator part of me stresses science as a process, tool, means and cultural artefact, not so much as a body of knowledge. So what about canoes? First, I'd like to make a terminological clarification. The Native in Richard's example is performing what I elsewhere called a "craft". This term was introduced by stipulative definition: stipulative definitions are useful because they fix terminology so that we can have some common way of talking about things. All the term means is that the plan or policy was not formed by grounding it on scientific research. Deciding when to use craft or when to use technology is the question here. So long as there are no other issues being debated, it is fine to stick with craft; the Native is also quite right to say that if another approach meant giving up something else he values he should decline. But another approach need not destroy her autonomy or anything else. It may; this is what I mean when I say that technology is morally committed, just as craft is. The craft as it is practiced (or at least as it is described by Richard) is an individual activity and thus does institute a value of autonomy. If this is the value that is to trump, say, productivity, that's fine. But there is no intrinsic reason why that should occur in every case. But also note that even if that were true, that (somehow every) science oriented approach were to do that it does not follow it should not be done. Richard is quite right to value autonomy on behalf of our Native friends. But this is only one competing value in many. The advantage of explicit proposals, once again, is so that debate and discussion can be fruitful (or even done). Suppose for example there were only so many birch trees to make canoes from. In fact this is the case; there are only a finite number, and a restricted number of those are suitable for canoe making. If a canoe is what is desired (a decision in itself) our friend might have to decide between the value of the forest and his autonomy. I for one hope that he is not confronted with that choice, but there are others, and the fact that I do not desire it either is another value judgement on my part. "No scientific way to break the deadlock" is thus in some sense true. In some cases there is, because there is a question of basic needs involved, and a scientifically aware ethics should be aware that some things (food, shelter, etc.) are more important to us in the sense we die or lead absolutely miserable lives without them. In the case of the canoe, perhaps this is not applicable. Ex hypothesi, it isn't. But there is another way to break the deadlock as well. Providing information so that people can make their choices and implement their values is in itself a way to break the deadlock, together with democratic approaches to government and economics etc. as we have discovered that these work especially well in that regard - and correspond better to our values that we seem to have largely in common. If we instead do not attempt to discover information in the best possible way we leave ourselves open to those who would either pass off bogus information or would use what they have discovered to bolster their own interests alone. This is where the "skewer" comes in. An environmentalist comes to the Native and says: "Stop cutting down the trees! You are destroying the forest!" (This example is not quite as far-fetched as it may seem. Raven (see the fourth part of the original series of documents) has told me about how Greenpeace's well meaning attempt to stop the seal hunts has produced all matter of misery amongst Inuit in Canada's north, who are now unable to pursue (to the extent that they would like) their traditional activity. Here autonomy and the importance of cultural traditions comes into conflict with environmental protection and animal rights considerations.) What is our friend to do? I am willing to hazard a guess that "Am not!" by itself is not a sufficient answer. A careful study of the matter, with the interests laid out based on this would help to resolve this conflict. (It might not: but not making use of the best available means, including the best knowledge of the systems in question and how to change them, to resolve it seems to be a recipe for disaster.)

It is my view that many of the ethical problems of technology can be partially mitigated by requiring technological proposals be public to some extent. Ethical review committees in hospitals, engineering firms, etc. are part of this process, albeit imperfect at the moment. (Technological literacy is vital for participating in these: one could be bamboozled either way on an issue if one does not understand the basis of what is being debated. But this is a problem for public education, of which more later.)

It is also important to realize that technology need not always replace crafts; it is again a public issue to decide where technology is needed and where it is not. However, when debating such matters one must remember that by leaving matters to non-scientific approaches one fails to gain the benefits of knowing a bit more about the world that they afford. This makes interventions more likely to succeed, and failure handled with more ease and sensitivity. If we know how a system operates, we are better able to fix it in case of failure. It is my view that this is part of the reason why non-scientific approaches to, e.g., health care, can be dangerous - not when they happen to work (e.g. by placebo, chance, careful observation and winnowing, keeping the patient occupied during spontaneous remission, etc.), but when they fail to, as then there is little ground to "clean up the mess". See the fourth paper in this series on indigenous knowledge for more on this and related matters.

Of course, not all ethical problems raised by technology are handled by these sorts of approaches. A broad technology ethics is not my intention here, only to sketch out some issues for further discussion. I believe what I have sketched applies to the three main strands of ethical thought. (It does not of course apply to ethical nihilism, but that was never the intention anyway.)

Finally, it should also be noted that I am not advocating technology as "big machinery", or anything of the sort. Even the smallest proposal to act informed by scientific research is technological in character, and so we must decide ourselves what sorts of technology we must have. The world is not in steady state, so it seems likely that new inventions will be needed to handle new circumstances. Keith Laidler has suggested that the time for what he calls empirical inventions - or what I have called crafts - is nearly over, as we have largely explored the part of the world that is accessible to us without scientific research. (I suspect that crafts will never quite disappear, just that new artefacts produced by craft alone will diminish - as has been already happening for at least the past 4 centuries.) Technology begins where crafts end just as science begins where common sense ends.

Ethics and technology: applied to the case of sociotechnology


When it comes to sociotechnology, there are two special features that make the ethical dimensions particularly problematic. One is that the social sciences being used are often underdeveloped. This results in technology being itself sketchy. It is not merely that the hypotheses used are more false than they could be, but also that they are less grounded and systematic. (See the first paper in the series for the virtues of scientific research as it pertains to social matters.) There is also some taint of ideology - which must be removed by careful research and "idea surgery" - in many of the social sciences, particularly economics. This leads to the second of the two features. Namely, that humans, being social animals, are "close to" social matters in a way they are not to atoms, galaxies and abstract algebra, and so may not be able to take one's proper distance from the matter they are studying.

That said, objectivity is compatible with partiality. If I really want to know what is wrong with my friend Raven, I'd better study neck anatomy (for that is where she is injured) objectively - I want to know what the neck is really like, not how it makes me feel or may appear to me. But I am being partial, in that I have selected a problem that is near and dear to me. This balancing act is especially difficult in social matters. Attempting to manipulate social systems is thus even more difficult.

Some people would argue that "scientific management" and such are intrinsically harmful and that we should steer clear of them altogether. A reason given for this is that planning people's lives is intrinsically authoritarian, authoritarianism is intrinsically bad, whence we should not plan. This is a bit too hasty. For one, it ignores the quite reasonable view that ethics itself is sociotechnological in character. If we postulate that human (or whatever else) flourishing is the subject matter of ethics, then sociology and other fields can be used to tell us what arrangements meet that goal the best. Ethics becomes then a technology. This does not escape the value judgements completely, but it at least shows how to ground them. All this is to say that planning, whether of economic arrangements, labour arrangements, power arrangements (i.e. politics), etc. that I am advocating need not be done from the "top down". (Nor necessarily from the bottom up, either. Studying the matter is what I am advocating.) Moreover, it does not follow that any particular arrangement ought to be universal. Perhaps we should arrange our buyers cooperative from the bottom, and also have universal direct participatory democracy in federal government - or the converse.

Why this matters


All this matters a great deal, because a social proposal to have a minimum of social planning is itself a plan. (Computer scientists find it handy to regard the empty case of a particular data structure to be a structure of the relevant sort. By analogy, the null plan of extreme right libertarianism should be regarded as a plan.)

It also matters because "[w]e have to create the future, or others will do it for us." (Susan Ivanova, Babylon 5: "Sleeping in Light".)

It is also important because it is my firm opinion that extreme technophobia (fear of technology) is as dangerous as extreme technophilia (love of technology). Why should we turn off our brains to the most effective ways we have discovered to get along in the world, especially as it doesn't mean losing our empathy and human concern in the process? Aristotle's dictum to seek the mean applies here too - and not globally, but on a case-by-case basis.

I mentioned above that deciding on technological proposals required technological (and hence scientific) literacy on the part of citizens. (It also requires some degree of economic and political democracy, of course, but that's another story for another time.) I suggested that this is partially a matter of education. I regard the study of education itself as partially a protosociotechology (and psychotechnology, more correctly, as it makes use of individual psychology as well as social psychology). Be that as it may, science and technology education is also a matter of the broad topic of science and technology policy. This topic I will address in the next paper in this series.


To return to the first page of the Science and Technology Policy Section, click here.

To email the SJC to comment, mail sjc@web.ca.

You may also click here to return to the SJC home page.