Could science fiction be used as a data source for educational research?

Heh heh, get it? It’s science fiction, it’s Data…

A fascinating and inspiring article came up in my Google Scholar search recently, Learning from Foundation: Asimov’s psychohistory and the limits of organization theory. In it, the authors, Nelson Phillips & Stelios Zyglidopoulos, argue for and demonstrate the use of science fiction literature as data in organizational research. I’ve often entertained the notion of using science fiction literature as data (not for organizational research, but rather educational development) but struggled with the question of how? Phillips & Zyglidopoulos describe an interesting approach, in which I see some pros and some cons, that has reignited my thinking on this matter, and given me some hope that I might actually figure out a way to do something similar with relevant data that I have collected.

Sci-fi schools: How are schools depicted in science fiction?

Readers who are familiar with this website of mine may be aware that for the past 15 years, or so, I have collected examples of depictions of schools and schooling in science fiction. I wrote about it once and described some of the examples in a brief article here on Education4site. Somewhat surprisingly to me, that article continues to be one of the most read here on Education4site.

Recently, I was asked if I would do a short seminar during the orientation week for new students at the University of Iceland’s School of Education (where I work). The idea with the seminar is to show new students “hidden sides” of faculty and administration in the college. The organizer was particularly interested in my futures work and thought that a seminar on that would be fun. But, everyone here at the college knows my futures work and those who don’t will quickly be exposed to it if they look me up. So, I suggested a seminar that would expose a more “hidden side” of me based on my collection of examples of how schools and schooling are depicted in science fiction.

I think the organizer was thoroughly gobsmacked. To her this was such a far out idea and, as such, very fitting for the occasion. But, from her reaction, I’m guessing that it wouldn’t have dawned on her the remote possibility that this could even be a thing.

Anyway, all of this got me thinking once again… So, I was looking over the material that I have and was once again, as so many times before, wondering what I should, or could, do with all of this sci-fi stuff that I’ve accumulated. And that’s what led me to discover the Phillips & Zyglidopoulos article.

Phillips & Zyglidopoulos argument for science fiction as a data source

Phillips’ & Zyglidopoulos’ primary concern is with the role of metanarratives in organizational research, especially the assumption that grand metanarratives can be developed to explain and predict organizational behaviour. They examine Asimov’s “psychohistory“, a central theme in his Foundation series, as an example of such a grand metanarrative, and, indeed, it might be said that it epitomizes the notion of a grand metanarrative.

Psychohistory is Asimov’s purely fictional creation of a science that quantifies social interaction to provide a predictive model. The model makes it possible to forecast social development over long periods of time with a level of precision that allows for detailed planning of events to promote a specific developmental trajectory. It is exactly such a prediction that is the backbone of the entire Foundation series’ storyline, when a social scientist uses psychohistory to predict the fall of a galactic empire, and its eventually rebirth. Of course, things go horribly wrong when unpredicted events occur. One of the primary issues that exacerbates the problem is how reliant the societies involved have become on the model’s predictive power.

Phillips & Zyglidopoulos treat Asimov’s Foundation series, and it’s science of psychohistory, as a hypothetical thought experiment, and, as such, a viable data source for serious research:

“In summary, Asimov’s Foundation trilogy can be thought of as a hypothesis regarding limits of social science.”

There is a weakness in defining science fiction as posing hypotheses that Phillips & Zyglidopoulos do not clearly outline. I think it would be safe to say that they are treating the Foundation series much like a scenario construction, as futurists have done for years. Scenario constructions can be viewed as hypotheses regarding long-term developments: they extrapolate from available evidence possible developmental paths and use the results to illustrate various alternative futures. However, there is a significant difference between a carefully researched and constructed scenario and a work of pure fiction.

Does science fiction hypothesize the future?

Carefully constructed scenarios are built on reproducible and verifiable empirical data regarding forces driving change (see more about this here). These data are building blocks that are independent of the individual, or individuals, engaging in the scenario construction process. Using the same methods and the same data, other individuals should arrive at the same, or very similar, scenarios. The same does not apply to science fiction.

The challenge for science fiction creators (whether they work with text, motion pictures, or whatever other medium) is not in creating stories that rest on a firm empirical foundation. Rather, it is to ensure that the components of the story appear plausible. Not only do they have a wide range of tools to use to accomplish this, but they also control all aspects of the narrative as it is being constructed. If they decide at some point to inject an element that is not entirely plausible, they can manipulate all underlying components to increase plausibility at any time and in any way that they choose.

In science fiction there is nothing that resembles a hypothesis in a strict sense. If we define a hypothesis as a proposal for which limited evidence exists that is explored for the sake of filling in evidentiary gaps then science fiction (or any fiction, for that matter) doesn’t cut it. Any apparent evidence can be manipulated by its creators in any way that they see fit before or after the case. For Asimov, the issues that arise concerning psychohistory are not derived from any existing, or even likely, evidence. He constructs a holistic scenario where all of the pieces are made to seem to fit together to produce a sense of plausibility. I think it would be safe to assume that most of Asimov’s effort was probably focused on figuring out how to make the predetermined storyline work rather than following any evidence, and I think that this would apply equally to other science fiction authors. I would like to say that the result is more of a pseudo-hypothesis than a scientific hypothesis, and that the difference between the two matters if we are engaging in scientific research.

To clarify my position here, a scientific hypothesis should take the form (=> is meant to suggest plausibility):

• If this, this, and this => that.

The result in the hypothesis equation (the “that”) is derived from the evidence that is available (the “this-es”), however scant it may be. Based on my experience and discussions that I’ve had with creative writers, I don’t think that this is how they work, especially not speculative science fiction writers like Asimov. The finished product in science fiction is, I think, more along these lines:

• If that => this, this, and this.

There is no presumed causal sequence that the result (i.e. the “that”) can be derived from. In fact, when the story is completed, after writing, rewriting, editing, etc., the result is probably more of a cause than anything else in the equation because the author has refined the story to fit the desired outcome. So, this is not a hypothesis. It is an alternative view of a possible reality manipulated so that all of the parts work together to make for a plausible whole.

My assumption about how creative writers work seems to sort-of fit what Asimov himself says about the origin of the Foundation series in this NPR interview from 1987, starting just before 14:00 (also see this page for more info about the radio show).

Is psychohistory a criticism of metanarratives?

Another point that I disagree with Phillips & Zyglidopoulos on has to do with their reading of the Foundation story. It may seem a minor point, but does, to some extent, enter into what they take the Foundation story to be evidence of. However, this might just be a matter of interpretation and they do mention the limitation that indeterminacy of interpretation poses for this kind of research, so we’ll bear in mind that we may just have read the story differently. That being said, my disagreement is about one of the central conclusions that they draw from the story and relates to the last of their “three important conditions that Asimov postulates for psychohistory to be predictive: that psychohistory doesn’t deal well with fundamental innovation.

In fact, as the Foundation storyline unfolds, psychohistory proves quite adept at predicting and dealing with all sorts of human innovation. Where it fails is at predicting circumstances that stem from non-human behaviour. The main antagonist in the story (the “Mule”) is not a “fundamental innovation” but, he is not a human either. He is a mutant. Psychohistory has been formulated with human behaviour in mind and the antagonist’s actions are contrary to what it was intended for. To apply psychohistory to the antagonist’s actions is kind of like trying to measure length in kilograms. Whether this is a limitation of psychohistory is debatable – it still seems to do reasonably well what it was intended to do, as long as we accept that its analytical and predictive power only extends to human behaviour.

Whatever we choose to conclude (remember, we’re talking about fiction – there’s a lot of wiggle room for alternative interpretations), in the radio interview with Asimov that I link to above, it would seem that despite psychohistory’s weaknesses that become apparent in the story, Asimov still believed, in 1987, that a predictive metanarrative, like psychohistory, was still plausible, if not feasible, even. This would seem to go against Phillips’ & Zyglidopoulos’ apparent interpretation of Foundation, and in particular psychohistory, as Asimov’s critique of metanarratives, or as they put it:

“…there are, as Asimov points out, powerful arguments against metanarratives in social science generally.”

I’m not sure that Asimov points this out at all. He points out certain limitations of our metanarratives that arise when we move into environments in which external unanticipated factors may impinge, but I don’t read him as arguing against metanarratives in particular.

So, what can science fiction as data tell us?

Big question for such a little blog, no? I’ve already described aspects of how I see science fiction in previous writings (see here for ex. Science fiction: Prediction or building blocks of the future), but I would hardly say that they provide definitive answers, nor do I see that as a task that I should be tackling. Science fiction, like any fiction, speaks to us on an individual level. It involves realities that are created in the head of the reader, observer, etc., not on the writers’ pages. These are not objective realities that can be prodded and tested. They’re more like little experiments that allow us to try various alternative futures on for fit. And it’s what we take from that that is the significant contribution that science fiction offers up. Some things will inspire us to action (bring about social change, invent new technologies), others will provide us with conceptual tools to broaden our discourse (new ways to talk about what can/should be).

For me, the conceptual tools and novel forms of discourse seem the most significant potential contributions that science fiction offers to social research. We see in science fiction possible perspectives that may not arise in our immediate, or experiential, environment, but that can enrich social activity and social development, and lead us down rewarding paths. Even when science fiction takes on its most pessimistic form (think for ex. 1984 or Farenheit 451), I think the takeaways are positive and work for the betterment of our societies. We at least have a better idea of, and ways to discuss, the things that we don’t want.

So, to get back to the original question, I think one of the roles that science fiction can play in research (in particular educational research) is to help identify things that affect the discursive environment. What tools do we have to describe relevant future settings and how are they used to construct imagined realities (i.e. how are the things that we are concerned with depicted in science fiction)? Conversely, we can consider what kinds of tools science fiction offers to broaden and advance discourse in our experiential environments, and how are they used.

I read Asimov’s Foundation series as attempting the latter. I did not see it as a critique of grand metanarratives, but rather as an expansion of metanarratives. To me, Asimov was saying,

“Look, you’ve created these wonderful models that can tell us a lot of valuable stuff. But, you limited your model to human activity. Look what can happen if non-human activity impinges on your model. Isn’t that something you might want to consider?”

And, indeed, it turns out to be a very relevant message today. Given the rapid advancements in AI, machine learning, and robotics, we may very well have to, in the near future, consider social development in societies with potentially significant non-human actors. This would likely call for radical changes in our models and metanarratives. Asimov has already provided us with some of the conceptual building blocks we will need to address those changes. Again, the question is what are those building blocks and how are we using them, if at all?

Have I found my new research agenda?

As I said at the beginning of this article, I find Phillips’ & Zyglidopoulos’ article inspiring. It has encouraged me to finally think seriously about what kind of data it is that I’ve collected over the years from my science fiction reading and what I can do with it in terms of research. I can’t say that I have everything worked out and can start tomorrow. My take on science fiction suggests a somewhat different approach than that applied by Phillips & Zyglidopoulos that I will have to clarify for myself before I can clarify it for anyone else. But, I’m optimistic and excited about the possibilities!