All in the Family. On Philosophy of Historiography and Philosophy of Science

Is historiography a science? When people ask this question, they are usually either intellectually curious about how different disciplines are related to each other or about the epistemological trustworthiness and intellectual respectfulness of

Is historiography a science? When people ask this question, they are usually either intellectually curious about how different disciplines are related to each other or about the epistemological trustworthiness and intellectual respectfulness of historiography. It is not clear, however, how these questions should be answered. We could give some intuitive answers or ask scientists and historians how they see the issue. However, this strategy would not work for three reasons. First, active researchers may not have a transparent picture of their own disciplines or they may not be able to conceptualize metalevel issues adequately. People might be stuck in some coarse-grained and old-fashioned philosophical ideas. We still hear discussions about falsificationism in science or lack of causality in historiography. One should not get upset over this observation. If there is something that science and historiography has taught us it is that things are not always as they seem to be, and people’s conceptions are often messy assemblies of various materials. This is also true about the nature of disciplines. Secondly, scientists and historians rarely have up to date knowledge about each other’s disciplines and their epistemological underpinnings. Asking them about the differences and similarities between disciplines would lead to a picture based on old-fashioned hunches. Thirdly, there are different types of sciences and different types of historiography. Different scientists and different historians would emphasize different aspects of their practice, and even if there existed some similarities that all the disciplines share, such weighted accounts would make them difficult to find.

It seems the question can be answered only by comparing philosophy of science and philosophy of historiography. We need to take the most systematic treatments of the nature of science and historiography and explicate the differences and similarities on the basis of these accounts. Sounds simple? It is not. There are three major problems in this approach.

First, the methodology of comparison is not clear. At first sight, it might seem that we can simply take some issue and then read the conclusions that philosophers of science and philosophers of history have drawn on the issue. For example, we could take an analysis of causal explanation in science and similar analysis in historiography and check the conclusions. However, the two sub-fields of philosophy have their own histories which have shaped their contents and the way things are expressed. For example, in the philosophy of historiography, analyses of causal explanation have been much scarcer than in the philosophy of science (due to the fact that there exists much more philosophy of science than philosophy of historiography). Some accounts of causal explanations have not been discussed in detail in the philosophy of historiography and it would therefore be premature to conclude that there are no causal explanations in historiography, even if influential philosophies of historiography said so. Even when philosophers of science and historiography seem to discuss the same issue, there might exist important differences in the ways in which the issue is framed. And this is of course true also within a single field of philosophy, as the whole point of philosophy is to discuss how certain issues, like causal explanation, should be approached. This leads us to the second problem.

The second problem is that philosophy of science is not complete and philosophy of historiography even less so. Of course, some progress has been made and we know, for example, that universal laws are not necessary for explanation in science or in historiography, and the absence of such laws is not what separates historiography from science, but more work needs to be done in both fields in order to compare science and historiography. In my view, our inability to compare historiography and science could be a symptom of the incompleteness of philosophy of science and philosophy of historiography. I think we can make the conjecture that there must exist similarities between historiography and science and therefore it is to be expected that a complete philosophical account of science (or historiography) will be able to cover both. There are three motivations for this conjecture:

First, it provides a methodological suggestion: We should attempt to apply accounts from the philosophy of science to the philosophy of historiography or the other way around (I think it still better to start from the philosophy of science as it has richer resources due to the sheer amount of research) and see how far we can clarify historiography or science with that account. Once we know which dimensions of historiography cannot be thus clarified, we should attempt to improve the philosophical account with respect to these dimensions. I discussed this strategy in the conference “The Role of Philosophy of Historiography” in Oulu 2017 and recently applied it (in a paper forthcoming in Journal of the Philosophy of History) to the case of causal explanations in historiography. In the paper, I argue that historiographical explanation can be understood along the lines of a general account of explanation (the so called interventionist account, taken from the philosophy of science) and, at the same time, it becomes possible to see how that general account can be understood from a different angle by examining historiographical explanations. Following the methodological suggestion and the strategy above, I had decent success in reframing issues in both philosophy of historiography and philosophy of science and in identifying core similarities between explanations in historiography and science.

Secondly, the conjecture explains the wide relevance of philosophy of historiography. If we assume that there exists a unified philosophical treatment of science and historiography, then philosophy of historiography gains relevance by being able to show how philosophical analysis of historiography can help to clarify issues that are present in many fields. Under the conjecture, philosophy of historiography is not a passive application of more general philosophical frameworks but a field that is equal in its contribution to the quality of those frameworks. A unified treatment of science and historiography is unlikely to come from philosophy of science alone. Building that treatment requires expertise in historiography.

Thirdly, science and historiography are often not separable. Sometimes a scientific explanation can be provided for a historiographical phenomenon and sometimes a historical explanation can be given for a scientific phenomenon. A chilling example comes from the Lake Bodom murders. In 1960, three people were murdered at Lake Bodom. The perpetrator was never identified and one explanation for this is that the crime scene was not recorded in detail and the police officers destroyed evidence with their actions. We know both (i) how quickly potential evidence is destroyed in those kinds of conditions and why, and (ii) why the police did not isolate the crime scene (the practice was not paid enough attention at the time). Science and historiography come together in our explanation of the mystery of Lake Bodom and it seems difficult to understand how such a unified narrative could be built if historiography and science did not share the same explanatory elements.

To sum up the latest points, it is difficult to compare philosophy of science and philosophy of historiography because these fields are incomplete. Moreover, the completion of the fields requires that we make some conjecture about their possible unification and I argued that we should conjecture that they can be unified.

The third problem in comparing philosophy of science and philosophy of historiography is that science and historiography are essentially normative practices in the sense that the questions “is this adequate?”, “does this have potential to epistemic achievements?”, “should we try that?” are omnipresent in both science and historiography. This means that philosophical accounts of science and historiography should have normative dimension in the sense that they should identify problems and limitations and suggest how those problems and limitations could be addressed. If philosophical accounts have a normative dimension, we cannot compare science and historiography through philosophy, as the philosophical accounts are not meant to be merely descriptive (even if the normativity is drawn from analyses of existing sciences, as it usually is these days). If we have an ideal of causal explanation, for example, that is not fully achieved even in sciences, then we should not expect that existing historiography achieves that ideal. We should not say that historiography is different from science because it does not achieve some ideals of science that science itself does not achieve.

A related problem is that science and historiography change. Philosophical accounts may at best be able to describe the current practices in these fields, but they must be constantly developed and updated. This indicates that philosophy of science and philosophy of historiography are essentially incomplete. The constant changes in science and historiography are also the main reason why I think the question “Is historiography a science?” is misleading. By locking down disciplinary identities and boundaries, we blind ourselves from fruitful interconnections between the current and future practices. Again, the question is not whether historiography should become a science – this would mean that there exists an unchanging monolith of science – but whether we should use all available epistemological means to understand the phenomena that are important to us.

Where does this take us? Perhaps one could already guess that I am against sui generis views of historiography, i.e. views that see a fundamental difference between what historiography and science study or how the investigation is conducted. As we saw, it is a good conjecture to assume that science and historiography share same elements. Moreover, sui generis views do not fit naturally together with the normative dimensions and (often parallel) changes in sciences. However, even if historiography is not sui generis, this does not mean that historiography is bad science or inferior in some other sense. Surely, natural sciences have many methods (experimental, for example, but we must remember that not all natural sciences are experimental) that historiography does not but this does not mean that historiography does not deserve its place in our epistemological practices. Our epistemology must be built around what we want to know and understand and history certainly is among those things. (Methods cannot define what is worthy of knowing. If they did, there would be no point in improving methodology and we would never be humbled by our methodological limitations.) Historiography is the best practice to achieve that knowledge and understanding. Surely, historiography can still be improved and probably we have to admit that historiography will never receive the same amount of evidential support as some sciences (due to the nature of the claims historiography makes or the scarcity of evidence) but this does not mean that there can be no knowledge of the past. I think it is better to view historiography as similar to other epistemic practices and admit its limitations than to argue that historiography follows its own logic and that the usual epistemic problems do not, therefore, apply to it. Such sui generis views avoid epistemic problems with an enormous cost as they make it difficult to see (i) what the relationships between historiographical knowledge and other forms of knowledge are, (ii) how historiography could be improved in parallel to other epistemic practices, (iii) how historical knowledge can be combined with other types of knowledge, and (iv) how accounts from the philosophy of science can make sense of historiography.

Finally, if there are no fundamental differences between historiography and science, there does not exist a fundamental difference between philosophy of historiography and philosophy of science. Based on my (admittedly limited) personal experience, it seems that some philosophers of historiography think that a philosophical account of some historiographical X is not an account of philosophy of historiography if the X is not unique to historiography. For example, as I followed my strategy of applying an (interventionist) account of explanation to historiographical explanation, I came to the conclusion that there seem to be similarities between historiographical and other explanations. Some scholars wondered how the resulting account can be an account of historiographical explanation if the account implies that historiographical explanations are similar to other explanations. In some sense I understand this reaction, but I think that a successful account of explanation in historiography simply is one that is able to capture and clarify explanation-related notions and issues that arise in historiography. It makes little sense to say that an account of explanation that is successful in these tasks is not an account of explanation in historiography because it is applicable to other fields as well. This is related to other reaction I noticed: Some people suggested that, if we are able to apply philosophy of science on historiography and see some patterns of similarities, then this is the main conclusion of the research. I do not think so. Philosophy of historiography should be seen as a field that analyses historiographical practices, inferences and conclusions. This is the interesting stuff: We want to understand historiography. It is a second-order question whether historiography and science are similar. If one day we understand both historiography and science, the real philosophical work is completed. Explicating the similarities and differences may then serve as our leisure activity.

(I thank Georg Gangl for useful comments on this text.)

Veli Virmajoki