Sohail Inayatullah (1998) has formulated a method of futures studies called Causal layered analysis (CLA). This method “is concerned less with predicting a particular future and more with opening up the present and past to create alternative futures” (815). It is a “method that reveals deep worldview committments [sic] behind surface phenomena” (815). CLA consist of analysis in/of four levels:

“The first level is the ‘litany’—quantitative trends, problems, often exaggerated, often used for political purposes.

The second level is concerned with social causes, including economic, cultural, political and historical factors.

The third deeper level is concerned with structure and the discourse/worldview that supports and legitimates it. [–] The task is to find deeper social, linguistic, cultural structures that are actor-invariant (not dependent on who are the actors).

The fourth layer of analysis is at the level of metaphor or myth. These are the deep stories, the collective archetypes, the unconscious dimensions of the problem or the paradox.” (820).

(It is important to notice that “myth” does not refer here to something that is untrue or even groundless. Rather, the term seems to underline that the contents in this level are not “regulated” by explicit efforts.)

There is much in common between philosophy of science and causal layered analysis. Let’s begin from the fourth level and make the discussion more serious step by step.

One could speculate that insights in philosophy of science arise from metaphors or “cut-feelings” or collective archetypes. For example, one could say that, after Kuhn had read Aristotle, he had a “feeling” that the ancient scholars really did see the world differently (see Kuhn 1987); or that, after meeting psychoanalysis (see Hacohen 2000), Popper gained an intuitive conviction that intellectual risks and errors are of utmost importance in our lives, that it is unproductive to shield against possible errors. One could also say that scientific realists are people who see scientists working (or at least talking) about all sorts of non-observable or theoretical entities and have the hunch that this work or talk must be about something out there. In this speculation, philosophy of science is seen as field which attempts to make explicit these gut-feelings or intuitions. This means that philosophy of science is already in the business of revealing and expressing “deep stories” that shape our understanding.

On another level, philosophical ideas can be seen as responses to existing myths. For example, Kuhn might be interpreted as responding to a myth about science as a cumulative and truth-approximating practice or, more sophisticatedly, to the myth that scientific changes, even major ones, can be neatly analyzed using only logical and conceptual tools. Popper can be also seen as reacting to a myth about scientific confirmation which says that evidence for a theory is the central driving force of science (Popper claimed that evidence against theory is much more important). Scientific realism can be seen as a response to the myth that empirical level of science is exhaustive, i.e. that science does not, in fact, have any content when it comes to questions of what world is really like. Realists deny that a description of science as a mere systematization of observations is an adequate one.

Notice that these philosophical views often react to previous views as “myths”. This implies two things. First, it seems that a myth does not have to be an “unconscious framework”. For example, Popper was reacting to the “verificationism” of logical empiricists (the thesis that meaningfulness and confirmation are linked) which was formulated many times in different forms (different formulations reflecting different views on confirmation). There was nothing unconscious about verificationism. However, the existence of these explicit formulations does not mean that verificationism was not based on a myth or a cut-feeling or that verificationism was not an embodiment of such myth. Actually, logical empiricists were never completely successful in their attempts to define verificationism and this can perhaps be seen as sign that they were trying to work out the details of a conviction that was never fully clear. If we attempt to understand the development of science on the basis of philosophy of science, it might be useful to see philosophical accounts both as theories and as myths, i.e. it might be useful to be faithful to their ability to clarify important issues and still keep one eye on the basic intuitive convictions that motivates those theories. In this way, we are able to be reflective on our current situation when it comes to our ideas about the (future) development of science.

Secondly, philosophy of science might be guilty of releasing myths that are counterproductive. My personal favorite is the view that, since Hempel wrote in 1942 that scientific explanations are based on laws and since there are no laws in the history, historiographical explanations are fundamentally different from scientific explanations. Even today this myth affects historiographical discourse, and if I were to estimate the possible futures of historiography, I would begin my scenario building from evaluating the viability of this myth. However, in defense of philosophy of science, it must be said that it is unfair to blame it for errors and interpretations that have a life of their own. Hempel’s views have been rejected for many decades already.

With respect to the third level of CLA, philosophy of science can be seen as analysis of structures that shape different aspects of science. For example, theories of explanation are not theories about any particular explanations or methods used in explanatory practices but about the nature of explanation: What it is that makes something an explanation, why explanations are important and how this is reflected on how explanations are sought and formulated, what makes a deep explanation etc. Answering these questions reveal implicit assumptions and motivations behind scientific practices that guide the actions of individuals and are therefore actor-invariant. These questions are important when we estimate the futures of science. For example, we might wonder to what extent sophisticated machine learning practices might provide explanations. Will they become uninterpretable black boxes or can they help us in our explanatory tasks? In order to answer, we need to define the threshold of explanatoriness and find ways to evaluate the depth of an explanation, i.e. we need theories of explanation. Moreover, philosophical accounts of explanation are often normative in the sense that they recognize “that causal and explanatory claims sometimes are confused, unclear, and ambiguous and suggests how these limitations might be addressed” (Woodward 2003, 7). This kind of normativity might play an important role when we evaluate which futures are desirable.

Another example from the third level are general theories of scientific development. These describe the general (macro) principles (or the lack of such principles, see Feyerabend 1975) that guide the development of science. For example, Lakatos (1970) argued that science consists of research programs that consists of series of theories sharing a hard core (untouchable part of the theory which is abandoned last in the face of problems) and of a protective belt of hypothesis that can be modified in the face of problems. A research program is progressive when a series of theories generate novel predictions (theoretical progress) and when the predictions are confirmed (empirical progress). We could say that research programs are structures that shape the work of a scientist and that the hard core and the shared positive heuristics (that tell what to do in the face of problems) justify those structures. It seems safe to assume that most of the time such programs are actor-invariant (one exception being Lavoisier’s work on oxygen). Interestingly, there is no clear cut point where we can say that a research program is no longer progressive enough and needs to be rejected. This seems to indicate that research programs are actor-invariant at least in the sense that one cannot deliver a clear knock-out blow for a research program. Given the protective belt and heuristics, we have to wait and see how the theory is perhaps improved. (We could say that, in science, the count to eight is of arbitrary length).

On the second level of CLA, we can perhaps classify philosophical works that are case-study oriented. Such works describe the methods, inferences and context of a scientific work. These philosophical analysis are not works of sociology, however, and we must understand Inayatullah’s reference to “social causes” in a very wide sense. This is not an ad hoc extension of CLA that is added in order show the connections between philosophy of science and CLA. For example, a genetic substructure may explain differences in the regional prevalence of coronary heart disease in Finns. Such explanation might be of interest in scenario building (in public health, for example) but it is “social” only if the category is understood in a wide sense.

For example, Siska De Baerdemaeker (in press) has recently described what they call “method driven experiments”. Method driven experiments are performed when scientists do not have enough knowledge about the target system to prefer some (set of) method(s) on the basis of that knowledge. In a method driven experiment, “the justification of the method-choice primarily appeals to what features the target would need to have in order for various established methods to be effective”, whereas in a “normal” target driven experiment, the method is chosen (perhaps even created?) on the basis of the features that the target system is known to have. Baerdemaeker’s work describes the role of method driven experiments in the search for dark matter and can be read as an explanation of the practices of that field (and Baerdemaeker’s work certainly has much to contribute on the “third level” (to use terminology from CLA) as an analysis of general inference-structures that shape experimental practices). Analysis of method driven experiments makes understandable how existing theoretical and methodological knowledge and opportunities affect what scientists do when the target system is a “mystery”. In this way, scientific work can be understood with reference to the immediate conditions surrounding the work (rather than subsuming the work under vague (macro) principles).

In a more theoretical level, we can note that there seems to be a connection between the second level of CLA and the so-called external vs. internalism debate (e-i-debate) in the philosophy of science. Let me explain.

As a first approximation, we can say that, traditionally, the factors that do not belong to the “internal” or “rational” workings of science have been understood as “external”. For example, Imre Lakatos argues that we should first attempt to “rationally reconstruct” the history of science as far as possible and only after such reconstruction use “external factors” to explain what could not be fitted into the rational reconstruction (1978, 102), and Larry Laudan argues that “application of cognitive sociology [external factors] to historical cases must await the prior results of application of methods of intellectual history to those cases” (1977, 208). It has never been quite clear what the e-i-debate is really about (see Shapin 1992) but the intuitive idea seems to be that the internal working of science consist of gathering evidence and evaluating theories on the basis of the evidence. There might be different views on the details of how these activities should proceed but in no cases should the decisions of scientists be affected by considerations external to theoretical merits and evidential relations. Then the e-i-debate would concern the question to what extent science can be explained without referring to external factors. What is the role of factors outside theoretical merits and evidential relations? Pure internalists would claim that science is never affected by external factors and pure externalists would claim that science is determined solely by external factors.

There does not exist pure internalists as far as I know [and I do not know the externalist camp well enough to judge whether pure externalists exist] but many philosophers have given external factors a secondary role in explanations of development of science. As we saw above, both Lakatos and Laudan argue that we must first provide “rational reconstruction” or “intellectual history” of science (i.e. an internal history of science) and only then investigate what residue must be explained by external factors. This means that we should provide a set of (macro) principles of scientific development that make the general structure of science understandable and focus on details later. Now, it is important to notice that, from the perspective of CLA, these (macro) principles seem to belong to the third level of analysis, as we saw above. If external factors are seen as causes that unsystematically affect science, then we could classify the study of external factors on the second level of CLA. This difference in the respective levels of internal and external factors would then make Lakatos’s and Laudan’s views on the superiority of internalist accounts understandable from the perspective of CLA. A “futurist-Lakatos” would say that it is more important to focus on internal histories because these give us deeper scenarios – nitpicking on unsystematic causes does not provide adequate basis for scenario building. Of course, the plausibility of this line of arguing depends on the assumption that rational internal accounts are, in fact, more systematic than explanations on the basis of external causes. If there does not exist any (rational) principles of scientific development then a scenario building based on those principles produces mere nonsense. I cannot take a stance on this issue here. It is enough that we note the connection between CLA and an old debate in philosophy of science. This might open fruitful lines of research.

I do not think there is anything important in philosophy of science with respect to the first level of CLA. However, we can note here that an important part of CLA is the movement between different levels:

“[A] dialogue between the different levels is sought. Interaction is critical here. By moving up and down levels and sideways through scenarios, [–] discourse/worldviews as well as metaphors and myths are enriched by these new empirical realities.” (Inayatullah 1998, 825.)

In philosophy of science there exists a long tradition of discussions about the relationship between case studies and philosophical theories. (See e.g. Donovan et al. 1988; JPH vol. 12 (2); Bolinska & Martin 2020). It might be highly useful to use insights from those discussions in CLA. For example, we might ask how the explanations in level 2 of CLA relate to the level 3. Are the explanations evidence for accounts in level 3 or is it the other way around, i.e. does the level 3 have to be in place before level 2 is even possible? Such questions are discussed in (methodology of) philosophy of science.

As a concluding remark, we can say that the similarities between philosophy of science and CLA indicates the potential value of philosophy of science in estimating the futures (of science). If CLA is a good approach in some cases of futures studies, so is philosophy of science. [Notice that we need to ask whether futures studies is a method or target driven epistemic practice (see above) but the answer must wait another essay.]

References:

Bolinska, Agnes & Martin, Joseph D. (2020). “Negotiating History: Contingency, Canonicity, and Case Studies”. Studies in History and Philosophy of Science Part A.

De Baerdemaeker, Siska (in press). “Method-Driven Experiments and the Search for Dark Matter”. Philosophy of Science.

Donovan, Arthur & Laudan, Larry & Rachel Laudan (1988). Scrutinizing Science: Empirical Studies of Scientific Change. Kluwer Academic Publishers.

Feyerabend, Paul K. (1975). Against Method: Outline of an Anarchistic Theory of Knowledge.

Hacohen. M. H. (2000). Karl Popper—The Formative Years, 1902–1945: Politics and Philosophy in Interwar Vienna, Cambridge: Cambridge University Press.

Inayatullah, Sohail (1998). “Causal Layered Analysis. Poststructuralism as Method”. Futures, Vol. 30, No. 8, pp. 815–829

JPH  = Journal of the Philosophy of History

Kuhn, Thomas (1987). “What are Scientific Revolutions?”, in Krüger, L. & Daston, L. & Heidelberger M. (eds.) The Probabilistic Revolution. Cambridge University Press: 7–22.

Lakatos, Imre (1970). “Falsification and the Methodology of Research program”. In Lakatos, Imre & Musgrave, Alan (eds.) Criticism and the Growth of Knowledge. Cambridge University Press. 91–197.

Lakatos, Imre (1978). The Methodology of Scientific Research Programmes. Cambridge University Press.

Laudan, Larry. (1977). Progress and its Problems: Toward a Theory of Scientific Growth. University of California Press.

Shapin, Steven (1992). “Discipline and bounding: The history and sociology of science as seen through the externalism-internalism debate”. History of Science 30: 333-369.

Woodward, James (2003). Making Things Happen.