Interpretivist Philosophy: Its Assumptions, Methods and the Criticisms

Interpretivist philosophy holds that the social world is not made of facts waiting to be measured but of meanings that people construct, and that research should therefore aim to understand those meanings from the inside and not to explain behaviour from the outside.

The position is usually introduced by contrast with positivism, and the contrast is genuine, but stating it as a preference for qualitative methods gets the order of reasoning backwards. Interpretivism is a set of assumptions about what exists and what can be known. Method follows from those assumptions; it does not define them.

The three assumptions, stated in the form a methodology chapter needs

Ontology: what exists. Reality is socially constructed and multiple. There is no single organisational culture waiting to be discovered; there are the accounts that different members give, which may conflict without any of them being wrong. The researcher does not arbitrate between them.

Epistemology: what can count as knowledge. Knowledge is produced through interaction between researcher and participant, not collected from a distance. Understanding is reached by interpretation, and interpretation is unavoidably shaped by what the interpreter brings.

Axiology: what part values play. The researcher's values are part of the study and cannot be bracketed out. Interpretivism does not treat this as contamination to be minimised; it treats it as a condition to be declared and reflected on.

The third is where methodology chapters most often fall short. Writing that the researcher will "remain objective" inside an interpretivist frame contradicts the position that was just adopted.

Where the position comes from

Weber's concept of Verstehen — interpretive understanding of social action in terms of the meaning the actor attaches to it — is the usual starting point, and Weber himself argued for combining it with causal explanation, not choosing between them.

Schütz developed the phenomenological grounding: social science constructs are second-order constructs, built on the first-order constructs that ordinary people already use to make sense of their world. The implication is that a researcher's categories have to be traceable back to the participants' own.

Geertz supplied the methodological vocabulary that has proved most durable, in the distinction between thin and thick description. A thin description records that an eyelid contracted; a thick description establishes whether it was a twitch, a wink, a parody of a wink, or a rehearsal — which requires knowing the code in which the gesture is read.

Three related labels are often used loosely alongside interpretivism, and distinguishing them is worth a mark: phenomenology studies lived experience as it is experienced; symbolic interactionism studies how meaning emerges through interaction and adjustment; hermeneutics studies the process of interpretation itself, with the hermeneutic circle describing how the parts and the whole are understood in terms of each other.

What follows for method

Because meaning is the object, instruments that fix the categories in advance are ruled out or heavily qualified. What remains is a family of approaches with a shared characteristic: the participant, not the researcher, supplies the terms.

Semi-structured and unstructured interviews dominate, because they let the participant introduce categories the researcher had not anticipated. Participant observation and ethnography go further, on the premise that what people do and what they report doing diverge in ways interviews cannot reach. Documentary and narrative analysis treat texts as artefacts of meaning-making and not as records of fact.

Sampling is purposive rather than probabilistic. The defence is not representativeness but richness: this case, these participants, because they are positioned to know. Theoretical sampling extends this, selecting the next case on the basis of what the analysis so far has made it necessary to find out.

Sample sizes are small and justified by saturation rather than by power calculation. The claim that saturation has been reached is contestable, and a chapter that asserts it without describing the point at which new data stopped producing new categories has not made the argument.

What interpretivist data actually looks like

A point that methodology chapters skip, and that examiners notice, is what the researcher ends up holding.

Interview transcripts run to several thousand words each, so fourteen interviews produce a corpus of perhaps eighty thousand words. This is not a smaller dataset than a survey's; it is a differently shaped one, and the labour sits in analysis, not in collection.

Coding that corpus means reading it several times, attaching labels to segments, grouping labels into candidate themes, testing those themes against the whole body of data, and discarding the ones that turn out to rest on two vivid quotations. The discarding is the part that separates analysis from illustration, and it is invisible in the finished chapter unless the writer says it happened.

Quotations in the findings serve as evidence and not decoration, which imposes a discipline: a theme supported by a single speaker is a single speaker's view, and presenting it as a finding overstates it. Stating how many participants a theme drew on, without pretending that the count is a statistic, is the honest middle course.

How quality is judged

Interpretivist work cannot be assessed on validity and reliability as a survey is, because it does not claim that a second researcher would produce identical results. Lincoln and Guba's alternative criteria are the standard substitutes.

Credibility replaces internal validity: are the findings a believable representation of the participants' meanings? Member checking, prolonged engagement and triangulation across sources support it.

Transferability replaces external validity: could these findings apply elsewhere? The researcher's obligation is to provide description rich enough that a reader can judge the fit to their own setting. The judgement belongs to the reader, which is the crucial move.

Dependability replaces reliability: is the process traceable? An audit trail of decisions, with their reasons, is the evidence.

Confirmability replaces objectivity: are the findings grounded in the data, not in the researcher's preferences? Reflexivity and the audit trail again supply the answer.

Using these criteria is itself a signal of methodological competence. A chapter that promises validity and reliability after adopting interpretivism has imported the wrong quality framework.

The criticisms, and which of them bite

Generalisation. The standard objection is that findings from fourteen interviews in two organisations cannot support claims about anything wider. The interpretivist reply is that statistical generalisation was never the aim and that analytical generalisation — to theory rather than to population — is what is claimed. This is a real answer, and it is weakened every time a study quietly generalises anyway in its conclusions.

Researcher bias. If interpretation is shaped by the interpreter, what distinguishes research from opinion? The answer lies in transparency rather than elimination: the audit trail, the reflexive account, and the presentation of enough raw material that a reader can see whether the interpretation is supported. Whether that is sufficient is a genuine and unresolved argument.

Replication. A second researcher would not obtain identical findings, and interpretivism accepts this. The objection has force only against claims interpretivism does not make, though it does constrain what the work can be used for.

Practical weight. Policymakers and managers frequently want a number, and a study that produces a rich account of six people's experience may be disregarded for reasons that have nothing to do with its quality. This is a criticism of the reception and not of the philosophy, and it is worth naming as such, not conceding it.

The positions next to it, which are frequently the better fit

Interpretivism and positivism are taught as a pair, and a great many business dissertations belong to neither.

Critical realism accepts that a reality exists independently of our knowledge of it, while holding that our access to it is always mediated by interpretation. It distinguishes three domains: the empirical, what is observed; the actual, what happens whether observed or not; and the real, the underlying structures and mechanisms that generate events. For research asking why something happens and not what it means, it offers explanatory ambition without the assumption that observation is neutral, and it is increasingly the position organisational research adopts.

Pragmatism treats the research question as deciding the philosophy rather than the reverse, and licenses mixed methods on the ground that the test of a method is whether it answers the question. It is the honest label for a great deal of applied business research, and it is frequently avoided because it can look like evasion. Declared explicitly, with reasons, it is not.

The practical implication is that a candidate who finds interpretivism uncomfortable — because the study does want to explain, or does want numbers alongside accounts — is usually a critical realist or a pragmatist who has not yet said so. Forcing the study into interpretivism because the data are interviews produces exactly the mismatch the philosophy section exists to prevent.

Declaring it in a methodology chapter

The competent version is short and follows one sequence: state the position, ground it in the research question rather than in preference, name what it rules out, and carry the consequence forward into the quality criteria and the limitations.

The sentence that does the most work is the one explaining why the question requires it. "This study adopts interpretivism because the object of study is how managers construe an ambiguous instruction, which has no existence independent of their construal" is an argument. "This study adopts interpretivism as it suits qualitative research" is a preference wearing the clothes of one.