Wednesday, 2 April 2014

Social science citations and literature reviews

In my role as Research Degree Tutor in the School of Health, I read quite a bit of social science based PhD research and have long been struck by how the use of citations differs from philosophy. I suspect that it is related to something else I have a problem with: the expectation that philosophy-based research should have something like a social science literature review. Perhaps even a systematic review. It is quite difficult to explain why not doing this in philosophy is not merely another example of the laziness of non-empirical research or another example of how such research is merely an ‘opinion piece’.

First here’s how social science citations look to me (forgive the mangled version of Harvard, for speed).

1: Meaning is use [Wittgenstein, L. (1953) Philosophical Investigations].

If one is a philosopher, this will seem a very odd sentence but I assume it won’t to social scientists. In philosophy, one simply doesn’t say that sort of thing even though Wittgenstein is an authority (the greatest philosopher of the C20?) and is widely credited with defending such a thesis. This example, although odd, is modelled on something which in another context is much more straight forward and which uses the citation to attempt to head off potential objections by an appropriate appeal to authority. For example in natural science one might say:

2: Smoking causes cancer. [U.S. Department of Health and Human Services (2010) How Tobacco Smoke Causes Disease: The Biology and Behavioral Basis for Smoking-Attributable Disease: A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health]

This seems fine but immediately suggests a problem for treating citation in this way in philosophy: conflicting authority is the norm rather than the exception. Returning to the subject area of sentence 1, we may also appeal to Davidson, for example, to say:

3: Meaning is not use. [Davidson, D. (2005) Truth, Language, and History. Oxford: Oxford University Press, p13]

So philosophy cannot rely on authority without running the risk of baldly saying odd things like:

4: Meaning is use [Wittgenstein etc] and is not use [Davidson etc].

The social science citations that bother me are those that move from a context where appeal to authority seems to me appropriate (like 2) to one where it is not (like 1). I read comments in methodology sections of social science PhDs along these lines:

5: There is no such thing as absolute truth. [Smithers 1998]

For every such authority-backed claim there will be a conflicting one somewhere out there in the world of published papers. So how can arbitrarily selecting one and citing it justify anything? But there is a further issue which this (and the foolish 1) prompts. Do we really understand what the sentence is saying? Sentence 1 fails as an assertion in a philosophy paper because it doesn’t tell us enough about the nature of meaning and of use and the way the connection sheds light on either. If Wittgenstein really does connect meaning and use, the connection is nuanced and stands in need of careful explanation. The work of articulating the connection, however, begins to display not only the nature of the thesis advanced but also its plausibility (cf this previous post). Once this is done, whilst a citation is necessary to avoid plagiarism, to show that it is Wittgenstein’s work that is being explicated, it is no longer asserting extra-textual authority. The text itself takes on that burden.  (This is a bit like moving from saying in the playground: 'My brother is bigger than yours' to saying 'My brother has taught me this technique with which I will now analyse your argument'.)

This difference between the way philosophy and social science relies on citation also connects to the absence of systematic literature reviews in the humanities. Reviewing the literature has a quite different status in the (social) sciences and the humanities.

I suspect that the background reason is that there is a presumption in the sciences that knowledge accumulates, that progress is made, and that this is reflected in the process of peer reviewed publication. Because of peer review, if something is published there is reason to hold that it is true unless there is specific reason (such as a conflicting publication) to doubt it. Hence also the habit of bald social science citation. If Smith has published the claim that p, then it is likely that p.

No such presumption operates in the humanities for at least two reasons. First, even though the humanities aim to tell the truth about their particular subject areas (one cannot just make things up), there is less idea of an actual accumulation of truth, perhaps even the very possibility of accumulating truth in areas like literary theory or the history specific events such as World War 1, eg. Second, the peer review process for journals establishes that papers meet a standard of rigour but that does not imply that they establish truth (even though that is their aim). The same journal might, without embarrassment, simultaneously publish conflicting papers, for example.

Hence there is no obvious point in undertaking a systematic review in the humanities because mere publication is no guide to anything. As a result the reviewer needs to take on some of the work of establishing what is probably true him- or herself. So there is no distinction between a literature review and the work of the research. But if so, reviews may as well be very selective.