The importance of the theoretical basis of Q-methodology

Abstract

This paper offers a commentary and response to ‘Beyond a set of procedures: Reclaiming the philosophical depth of Q-methodology’ by Susan Ramlo. In agreement with Susan Ramlo, an account is given of how important it was to William Stephenson to focus not just on what he called Q-technique but also on Q-methodology, which includes the set of theoretical principles informing the technique. For Stephenson, Q-methodology was contrasted with the ‘R-methodology’ associated with Pearson, Spearman, Burt and the other members of the London School who were centrally involved in developing the technical procedures of correlation and factor analysis. Stephenson was in the ‘belly of the beast’ in the sense that he worked with this group but strove to critique and re-articulate their methodological assumptions. This historical context acquires a fresh importance once we factor in that the London factorists were central figures in the development and promotion of eugenics.

Q-technique and Q-methodology contrasted with R-technique and R-methodology

The paper by Susan Ramlo offers a kind of eulogy to Q-methodology. It makes a spirited case for foregrounding the more philosophical and theoretical aspects of Stephenson’s methodological innovation, including the ambition to define and study ‘subjectivity’ and the novel approach to scientific epistemology and rigour that follows. The latter includes the questioning of objective detachment, the reframing of the problem of scientific generalisation (glossed in an interesting way by Ramlo as ‘substantive generalisation’, and appropriately related to abduction), and the incorporation of ‘quantum’ notions like complementarity.

I completely agree that Q-methodology needs to be approached not simply in terms of its procedural or technical aspects but also in terms of its more theoretical aspects. I won’t repeat those theoretical aspects here since I have put these arguments in print on several occasions as part of my effort to contribute both technically and theoretically to the development of Q-methodology (Stenner, 2009, 2022; Stenner & Capdevila, 2019; Stenner & Stainton Rogers, 2004; Stenner et al., 2025; Watts & Stenner, 2012). Since Stephenson himself was also very clear on this ever since he first developed his methodology in the 1930s at University College, London (see Stephenson, 1936a, 1936b), it is valuable to consider his own views more concretely. When he moved to UCL to work with Charles Spearman and his circle, Stephenson quickly became aware of the limitations of factor analysis and, himself an able mathematician and trained physicist, began pursuing his own path on both technical and theoretical fronts. As Ramlo notes, this led to an interesting disagreement with Cyril Burt which the two of them played out, for the record, in print (Burt & Stephenson, 1939). The disagreement was precisely about the importance of the broad theoretical frame that is needed if the data matrix prepared for factoring is ‘inverted’, as is the case with Q-methodology (Stephenson, 1936a). I will not repeat those arguments here (see Brown, 1980) other than to say that Burt could see no significant theoretical issues at play in treating the variables of a data matrix as cases and the cases as variables, whilst Stephenson (rightly in my view) was very sensitive to the relevance of the theoretical assumptions at play (see Watts & Stenner, 2007).

The point I am making here is simply that, from the very birth of Q-methodology, for Stephenson the technical issues that interested him (reversing the rows and columns in a set of correlational data to be factored) and the theoretical issues (the theoretical meaning of the data and the results) are related and must be considered part of a coherent methodology. This remained a consistent theme throughout Stephenson’s life. We find a particularly sustained treatment of this theme in his book The study of behaviour (Stephenson, 1953), written when he was a member of the Department of Psychology at the University of Chicago. In the introductory chapter of this book, Stephenson states that his “concern is with far more than the simple operations called ‘Q-technique’. Rather, it is with a comprehensive approach to the study of behaviour, where man [sic] is at issue as a total thinking and behaving being” (p. 7). ‘Q-technique’ is thus clearly differentiated as the purely technical part of the whole that is ‘Q-methodology’. The study of behaviour is, unashamedly, a campaign on behalf of the latter. ‘Q-methodology’ is thus a combination of ‘Q-technique’ with a whole set of theoretical principles. As Stephenson puts it, he claims “no original authorship for many of the principles themselves, but the methodology as a whole, we hope, has some newness” (p. 7).

Q-methodology is contrasted throughout with ‘R-methodology’ but also ‘Q-technique’ is contrasted with ‘R-technique.’ In using the designation ‘R,’ Stephenson was referring to Karl Pearson’s correlation coefficient (r). For Stephenson ‘R-technique’ involves the correlation of tests (IQ, reaction time, etc.) as part of the study of so-called individual differences (I say ‘so-called’ because in fact the study of individual differences required the correlation of data across a large number of individuals). ‘Q-technique’ by contrast was based on the correlation of what Stephenson called ‘single cases.’ By this, he did not mean that only one person at a time can be studied (though he often followed this procedure) but that correlations were to be computed between whole Q-sorts (each Q-sort thus being a ‘single case’). In making this distinction, Stephenson is already drawing attention to important theoretical differences at play in the technical matter of correlating data. The mathematical formulae may be identical in both cases of correlation (by test score and by Q-sort), but the very data upon which they operate must, Stephenson insisted, be clearly differentiated. Inverting a matrix of R data from tests for factoring is logical nonsense but it is perfectly appropriate with Q-sort data.

Pearsonian methodology as implicit and problematic

I have rapidly summarised the basis upon which Stephenson ‘began to object’ against the (R) methodology assumed by the London School of factorists. It is important to note, however, that Stephenson did not think that R-methodologists were aware of being R-methodologists. Rather, they assumed this methodology uncritically. So, in coining the distinction between R and Q, Stephenson was drawing explicit attention to the fact that science always involves methodology, whether explicit or implicit. Stephenson states very clearly that the implicit methodology which he rejected “stems . . . from Karl Pearson’s Grammar of Science” (p. 11). This book, he suggests, expresses the “underlying belief of the traditional factorist.” Stephenson summarises this belief as follows:

the whole wide universe, the geography of the earth, its flora and fauna, and every human action, were grasped by Pearson as a vast matrix, cemented together inexorably by correlation . . . this correlation has only to be studied in its own right in order to reach great scientific truths. (p. 11)

The certainty concerning the value of R-methodology on the part of Pearson stems from the belief that the scientist need only grasp the given correlational cement of the universe by statistical inference in order to describe things as they really exist. Stephenson’s Q-methodology is designed to forego these conceptions and indeed to challenge the entire nexus of protopostulations, including the “corresponding belief in abilities, capacities, potentialities, and similar general propositions . . . about individual differences.” In proposing the name ‘R-methodology,’ Stephenson is doing nothing less than identifying as deeply problematic the belief system informing the work of the London School of factorists, a set of underlying beliefs “so deeply ingrained in the factorist’s thinking that it takes a very special effort on his part for him to become aware of them” (p. 12). These beliefs “constitute a complex system or, as we now call it, a methodology. We call it ‘R-methodology’.”

Stephenson could not be more explicit about his target. But he is also clear that his alternative, namely Q-methodology, is an effort both to be as up-front as possible about the beliefs informing his science and to open the resulting principles to scrutiny. Given the considerable shaping influence of R-methodology on the discipline of psychology as a whole, Q-methodology was to pose quite a challenge to the discipline. Stephenson did not shirk this challenge, stating on the very first page of The study of behaviour that his concern is not with Q-technique alone but “with a challenge to psychology, in certain of its aspects, to put its house in scientific order.” Q methododology, he states, includes a philosophy of science that embraces all logical analysis “except such as have led to the excess of reductionism,” (p.1) and that distinguishes between general and singular propositions when proofs are at issue. R-methodology is found guilty both of reductionism and of ignoring singular situations in favour of solely general propositions.

R-methodology and eugenics

Stephenson, perhaps diplomatically, largely ignored in his written work the politics lurking in the background of R-methodology (his reflections on the Burt affair in Stephenson, 1976, discussed below, are an exception). Today we can no longer ignore the fact that this London School against which Stephenson wrote was the very epicentre of eugenics. In January 2021, UCL finally apologised for the role it played in developing and legitimating eugenics which “provided justification for some of the most appalling crimes in human history: genocide, forced euthanasia, colonialism and other forms of mass murder and oppression based on racial and ableist hierarchy” (UCL, 2021). UCL ‘de-named’ several spaces on its campus named after prominent eugenicists like Francis Galton and Karl Pearson. In March 2021 the BPS announced the ‘retirement’ of their prestigious Spearman Medal, the BPS Research Board noting his ‘complex legacy’ (BPS Updates, 2021).

In this respect, Stephenson’s stand against the kind of scientific belief system articulated by Pearson must also be placed in the context of Pearson’s political belief system. Naturally, Pearson’s scientific and political belief systems were tightly interconnected. Hence in his book National life from the standpoint of science (Pearson, 1901, p. 41) Pearson describes his politics as based on a “scientific view of the nation.” The nation, defined from his supposedly scientifically objective perspective, is “an organised whole in continual struggle with other nations, whether by force of arms or by force of trade.” This struggle, he proposes, is the source of all progress and it is a struggle between races. The supposed “high state of civilisation” is therefore the product of just one inevitable process: the violent “struggle of race with race” (p. 42). If a nation is to survive a crisis or to “maintain its position in this struggle, it must be fully provided with trained brains in every department of national activity, from the government to the factory,” and for this it needs “a reserve of brain and physique” (p. 42). Guided by science, eugenicists aimed to lend a helping hand to the higher races in their struggle against the lower and in so doing helped the nation keep its dominant rank in the global hierarchy of power. Hence it was of particular importance to Pearson that the citizens of a nation be trained in science. To strengthen and energise the nation, what is needed is a great increase of powers of observation, of scientific reasoning, and of “knowledge of scientific method” (p. 40).

Eugenics was of course founded by Galton who funded the Chair of eugenics at UCL that was taken up first by Pearson and then by Ronald Fisher. In 1904 Pearson wrote to Galton saying:

I think immense good can be done by a careful statistical study of what tends to deteriorate and to strengthen a nation physically and mentally from the standpoint of the individuals from which it is reproduced. I hope in this sense that Eugenics covers not only a theory of better breeding, but a study also of Kakogenics or of the bad breedings current at present. (Letter from Pearson to Galton from 1904, cited in Norton, 1978, p. 103)

So when Stephenson took aim at the R-methodological belief in abilities, capacities, and other ‘individual differences,’ we ought not avoid the fact that there was also a deeply problematic political dimension informing the scientific beliefs he was critiquing and reframing. It is interesting in this respect to note that when Mendelian genetics was introduced to the UK by William Bateson (who coined the term ‘genetics’ in 1905), it sparked a controversy known as the ‘biometric-Mendelian debate.’ An anonymous participant in this heated debate (possibly Bateson himself) ironically described the Biometric School of the eugenicists as “organised along the lines of a field army,” with Pearson a supreme “Field-Marshal” imposing severe discipline to prepare his troops “to apply mathematical methods to any problem.” Mostly, writes the anonymous writer, the result of these efforts was to obscure “the geography of the country of its invasion by the smoke of battle, produced by the burning of its ‘correlation’ gunpowder” (see Norton, p. 168).

I noted above that Stephenson did offer some reflections on the Burt affair. Sir Cyril Burt turned out not to be such a paragon of science because he was found to have faked his twin studies data and to have invented some missing assistants to cover his tracks (Willmott, 1977). Stephenson (1976, p. 114) notes the:

. . . use of factor-analysis in the framework of “the psychology of individual differences”; the matter was (and remains) of paradigmatic proportions, within what we can justly call “objective-positivist” methodology. Tens of thousands of studies have been pursued and thousands of practitioners are now engaged in mental testing based on this paradigm, in almost every form of applied psychology (clinical, educational, social, political or whatever, wherever abilities, personality, and social influences are at issue). In this context Burt was indomitable . . .

Burt was doggedly committed to Galton’s programme for identifying talent and character and their opposites, but Stephenson notes that already by 1931 he questioned Burt’s premises about mental testing and intelligence. Stephenson confesses to have constructed the tests used by the Royal Air Force in World War II, but states that, unlike Burt, he gave them no theoretical importance. Indeed, Stephenson objected to the 1944 Education Act in an article forthrightly critical of testing school children and using results to select for Grammar Schools (a position that was ignored). As he puts it: “I was therefore far from being merely ‘theoretical’ about Burt’s influence. I did what I could to offset it – but of course it was impossible to challenge Burt’s pre-eminence in this educational matter” (Stephenson, 1976, p. 117).

In this account, Stephenson states that his development of Q-methodology in the early to mid-1930s was born from his growing dissatisfaction with the premises assumed by ‘the mental test methodology’ upon which the work of Burt and Spearman depended for its scientific integrity. At the risk of repetition, it is worth quoting the following rather clear statement:

I used the term “methodology” very deliberately, to mark the beginning of what has now to be regarded as a new paradigm, in which “inherent-relatedness” and the “single case” are axiomatic, as distinct from “objective-positivism” and individual differences for samples of persons. Subjectivity was my aim, the rightful objective of pure psychology and not objectivity and the biological nexus. (Stephenson, 1976, p. 118)

In this respect it is notable that Stephenson, from the mid-1930s, was quite explicitly articulating a new paradigm for psychology as what he called a subject science, orientated towards the standpoint of the subject (a phrase he uses more than once in his 1953 book).

Concluding thoughts

I have tried to outline an important part of the context in which Stephenson strove to both acknowledge the study of ‘subjectivity’ and put it on the scientific agenda of psychology. In questioning the objective detachment of people like Pearson and Burt he was also calling into question the very possibility of objective detachment and championing a more reflexive approach to knowledge. These points are of course quite consistent with the argument made by Susan Ramlo. Perhaps I differ from her position, however, in that I do not think that the theoretical issues articulated by Stephenson in the name of Q-methodology belong to philosophy rather than science. In championing Gary Gutting’s argument about What philosophy can do (Gutting, 2015), the suggestion seems to be that the credit for Stephenson’s insights is to be laid at the door of philosophy, not psychology.

I do not wish to deny the real-world relevance of philosophy as a pragmatic discipline that can help people in many fields of general interest to clarify their assumptions and improve their logic. Philosophy can indeed help people to discover and reflect upon their own beliefs and to recognise these as perspectives amongst other perspectives on issues of general relevance. I disagree, however, with Gutting’s view that scientific methods cannot address questions of meaning and value, or questions of consciousness, morality, and feelings like happiness. Gutting argues that these problems belong to philosophy and that the empirical methods of science cannot address them. Stephenson, however, was not a philosopher and did not see his methodology as a contribution to philosophy, and although Q-methodology does both draw upon and carry interesting implications for philosophy of science, it is a psychological methodology designed precisely to study problems related to meaning and value. Stephenson would have completely disagreed with the proposition that psychology cannot address subjectivity, and throughout his career he rightly took great pride in the technical aspects of his method, including his invention of the Q-sort as a technical means of data collection which elegantly allowed him to correlate and factor individual Q-sorts (this being the basis of his notion of ‘single cases’). Furthermore, were it not for his training in the mathematical basis of physics it is unlikely that Stephenson would have been able to accomplish these technical, procedural innovations.

In sum, I thank Susan Ramlo for directing attention to Q-methodology as a phenomenon of significant interest both for philosophy and for psychological theory, research, and practice. I hope that ‘Beyond a set of procedures’ draws interest both to the technical and the theoretical aspects of Q and other comparable methodologies, and opens some fresh questions about the relations between philosophy and psychology. A theme I find particularly fruitful for future research here is the relationship between Q-methodology and the sort of process ontology of experience articulated by thinkers who directly inspired Stephenson, including William James and A. N. Whitehead (I made a start on this in Stenner, 2009). Stephenson’s concept of subjectivity was the product of a complex engagement with process-orientated sciences and philosophies, including a critically reconfigured behaviourism. His notion of subjectivity would, I suspect, be as unfamiliar to many philosophers as it is to many psychologists. Subjectivity, for Stephenson, was a set of real processes: the activities of thinking, dreaming, imagining, feeling, for example, are no less real for being unobservable to a third party. This is why he was keen to insist that there is no ghostlike realm of mind distinct from that of the body.