Praxeological Analysis: A New Qualitative Methodology

Abstract

This paper introduces Praxeological Analysis (PA), a new qualitative methodology for investigating psychological phenomena by examining their situated sense within interaction and talk. PA draws upon and develops ideas from three intellectual resources: (i) praxeology, (ii) gestalt psychology, and (iii) the method of grammatical investigation found in the investigative ordinary language philosophy of Ludwig Wittgenstein. We integrate these three resources through the concept of linguistic gestalts, which serves as the foundation for PA. We argue that once psychological phenomena are respecified in terms of linguistic gestalts, the appropriate mode of investigation is what Wittgenstein termed grammatical. PA implements this mode of investigation by generating realizations through descriptions of actual, imagined, and fictionalized situations. The goal of such investigations is to explore the grammar—the practice-embedded rules for meaningful use—of psychological terms, as this reveals their real-world significance. In PA, we develop this insight into a structured methodology that qualitative researchers can systematically apply. By outlining the conceptual foundations of PA and demonstrating its methodological application, we position PA as a novel and transformative approach in qualitative research.

Introduction

In this paper we introduce Praxeological Analysis (PA), a new qualitative methodology for investigating psychological phenomena by examining their situated meaning within interaction and talk. PA draws upon and develops ideas inherent to three intellectual resources: (i) praxeology, (ii) gestalt psychology (Koffka, 1935), and (iii) the method of grammatical investigation found in the investigative ordinary language philosophy of Ludwig Wittgenstein (1953/2009) and Frank Ebersole (2001b). ¹ We combine these three resources in the idea of linguistic gestalts. We argue that once psychological phenomena are respecified in terms of linguistic gestalts, the appropriate mode of investigation is what Wittgenstein termed grammatical. PA implements this mode of investigation by generating realizations through descriptions of actual, imagined, and fictionalized situations. The aim of such investigations is to elucidate the ‘grammar’ of psychological terms—the practice-embedded rules that shape their meaningful use. We engage in such investigations because, as Wittgenstein put it, “Grammar tells us what kind of object anything is” (1953/2009, Section 373). We seek to develop this insight into a structured methodology through which grammatical investigation can be implemented by qualitative researchers to elucidate phenomena. In what follows, we first outline the conceptual foundations of PA, its central characteristics, and its relation to and differences from existing approaches. We then explicate how to conduct a PA, before discussing how PA can benefit psychology as a discipline.

The Conceptual Foundations of Praxeological Analysis

Praxeology

Praxeology is the study of human action and practices, focusing on the structures and processes through which actions are performed and understood. The term derives from the Greek ‘praxis’ (action, practice) and ‘logos’ (study, reason, or logic). While the early to mid twentieth century saw a number of theories of praxeology, e.g. Slutskii (1926/2004), Kotarbinski (1965), and von Mises (1949/1998), ² we neither invoke nor draw upon these theorists and nor do we seek to provide a theory. We use the term ‘praxeology’ and its derivatives to denote the situated study of action, interaction, and co-action as that features in and is constitutive of the production of meaningful situations.

Gestalt Psychology

Gestalt psychology emerged in the early 20th century as a broadly phenomenological approach rooted in the holistic perspectives of philosophical psychologists such as Franz Brentano, Carl Stumpf and Ernst Mach. Central to the development of gestalt psychology is the notion of gestaltqualität, coined by the German philosopher Christian von Ehrenfels (1890/1937), to denote a quality that emerges from how an object is assembled but which is not reducible to the sum of its parts taken as discrete elements. This quality was conceived by the Berlin School of gestalt psychology as a more primary gestalten, or structured whole (Isaac & Ward, 2019). Grounded in this idea, gestalt psychologists rejected the widespread assumption that the aim of psychology should be to break psychological phenomena down into basic elements or atoms, instead foregrounding the mutual dependency of parts and wholes. ³

Among the key distinctions in Gestalt psychology is the contrast between molecular and molar analysis. Molecular analysis examines the smallest components of a phenomenon—atoms or elements—while molar analysis focuses on wholes, systems, and unified structures. In other words, molar analysis examines gestalts (integrated and meaningful wholes) rather than dissecting phenomena into their smallest parts. To put it simply: molecular analysis focuses on atoms or molecules (the metaphorical building blocks). Molar analysis focuses on the entire organism, event, or structure as a cohesive whole.

Perceptual Gestalts

Beyond psychology, Gestaltism is perhaps most widely known for the famous Gestalt pictures, such as the duck-rabbit, faces-vase, and so on. While there are many observations that can be made with reference to these gestalt pictures, we here focus on the mutual dependency of parts and wholes, or what is called an internal or meaning relation between constituent and whole. Consider Jastrow’s duck-rabbit (see Figure 1) and the Ruben faces-vase (see Figure 2).

Figure 1.

Jastrow Duck-Rabbit.

Figure 2.

Ruben Faces-Vase.

Both these pictures demonstrate that the parts of the images have the meaning they do only in so much as they are experienced as parts of the whole object; they are what Wolfgang Köhler referred to as ‘dependent part qualities’ (1969/2015, p. 54). For example, the long protrusion of the duck-rabbit is identifiable as ears if you see them as the ears of the rabbit, rather than as the bill of the duck. Relatedly, you cannot arrive at the description of the whole, for example “this is a duck” or “this is a vase”, based purely on a molecular analysis of separate parts taken as discrete elements which are unrelated to the meaning of the whole. You must see the putative ‘elements’ as features, constituents, or regions of the whole through a molar analysis. Referring to the parts of the whole as ‘features’, ‘constituents’ or ‘regions’ (and not as ‘elements’) serves to highlight the way in which parts of the whole are seen as intrinsically or internally related to the whole. As noted, Köhler would refer to these constituents as “dependent part qualities”; the philosopher Aron Gurwitsch (1953/2010) and the sociologist Harold Garfinkel (1967) would refer to these constituents as members to emphasise the same point (think of the term dismembered for the removed or disconnected region of the body).

Behavioural and Praxeological Gestalts

Gestalt psychologists, and those who inherited their critique, took the idea of the mutual dependency of parts and wholes and considered it not just in terms of perception, but also in terms of behaviour and action. For example, in proposing a more inclusive notion of behaviour, Kurt Koffka (1935) adopted Edward Tolman’s (1932) terms ‘molecular’ and ‘molar’ behaviour to distinguish between two levels of behavioural analysis. ⁴ Molecular behaviour refers to discrete, momentary responses to stimuli that occur in a linear sequence, while molar behaviour encompasses the broader, purposeful actions experienced in everyday life. Koffka argued that psychology should focus on molar behaviour, emphasizing that its defining characteristic is its occurrence within a behavioural environment.

Consider an example: the actions of mourners at a funeral—such as bowing heads, maintaining solemn expressions, and participating in ritualized gestures—are purposeful behaviours that reflect and sustain the collective grieving process. These actions are meaningful not as isolated responses to stimuli but as contributions to the broader social and emotional environment of the funeral.

A few decades after Koffka wrote on behavioral gestalts, the sociologist and founder of ethnomethodology, Harold Garfinkel, developed the idea of praxeological gestalts as a core component of ethnomethodology. ⁵ Although sharing some similarities to Koffka’s account of molar behavior and behavioral gestalts, Garfinkel’s account is significantly richer and developed in more detail. Garfinkel’s praxeological gestalts differ in one key feature: the context is praxeologically produced and reflexively made witnessable (or accountable) by the participants, (or members) and Garfinkel refers to this as the gestalt contexture. Talk of gestalt contexture here emphasizes that the context is both actively produced through inter- and co-action and that the relationship between gestalt whole (situation/social phenomenon) and the actions of members that are productive of it, is reciprocal: the identity of the situation and that of the constitutive actions is internally related. Garfinkel proposed that members, through their methodic actions, create and make visible the gestalt whole, while the whole gives meaning to the members’ actions in a reciprocal process. So, where in Koffka’s behavioral gestalt, the environment which provided the context or gestalt was a passive context, serving a kind of framing role, in Garfinkel the situation is actively, praxeologically, produced ongoingly by the constituents or members.

Consider again the example of a funeral, something which is recognizable as a phenomenon of social order and for which we have a term: funeral. We can recognize and talk about funerals because they have recognizable form. When we observe a funeral we see individuals enacting shared taken-for-granted methods to both constitute (produce) and render visible (accountable) a shared social and emotional order. Through their methodic actions—such as maintaining solemn expressions, engaging in moments of silence, or following ritualized gestures like placing flowers or bowing their heads—participants simultaneously enact mourning practices while making their grief recognizable to others. This mutual orientation produces the funeral as a witnessable social unit, functioning as a praxeological gestalt: a holistic structure that shapes and is shaped by the actions and identities of its members doing ‘mourning.’ Reflexivity is evident in how participants adjust their expressions and behaviours in response to the social and emotional atmosphere, ensuring that their actions align with and reinforce the collective sense of mourning.

Grammatical Investigation, Linguistic Gestalts, and Investigative Ordinary Language Philosophy

Building on our discussion of perceptual, behavioural, and praxeological gestalts, we now extend our discussion to what we term linguistic gestalts. The idea of linguistic gestalts is derived from the work of the philosopher Ludwig Wittgenstein (1953/2009), and is another term for what Wittgenstein called language games. We favour the term linguistic gestalts because it emphasises the continuity with other gestalt phenomena. In this section, we will begin by briefly explicating Wittgenstein’s notion of a grammatical investigation, before saying a little more about language games.

Grammatical investigation is a distinctive elucidatory method developed by Wittgenstein in his later work. Rather than seeking traditional philosophical definitions or constructing theories, grammatical investigation examines how words are used within diverse contexts of human activity. Wittgenstein’s method contrasts sharply with approaches that attempt to solve philosophical problems through theoretical explanations; instead, it aims to dissolve such problems by revealing how they arise from misleading pictures ⁶ of language and mistaken assumptions about meaning. Through careful attention to the everyday use of language, we can free ourselves from the grip of these misleading pictures—misconceptions that frequently generate confusion in philosophy, psychology, and other disciplines. In this way, grammatical investigation reminds us that words gain their significance from their role within our practices and forms of life. ⁷

Wittgenstein introduced the concept of a language game to help us recognize key features of language use—especially the reciprocal relationship between what is said and the specific situations in which speech occurs. Language games function as structured wholes, or linguistic gestalts, meaning that words and expressions acquire significance through their relation to the activities of which they are a part, rather than through abstract definitions or fixed meanings.

In addition to highlighting language as practical and contextual, Wittgenstein’s use of the game metaphor in the term language game draws attention to two further points.

The first relates to definitions. Wittgenstein uses the term ‘game’ to illustrate that the things we group under a single word don’t always share one essential feature. Games like chess, hide-and-seek, snooker, rugby, and catch do not have a single trait in common. Instead, they are related by a series of overlapping similarities—what Wittgenstein famously calls “family resemblances.” Wittgenstein uses the term language game to emphasize a similar point about language: the meanings of words are not captured by universal definitions but by how they overlap and interconnect within specific contexts.

The second relates to meaning and context. Understanding the meaning of actions within a game requires understanding both the rules of the game and the broader context in which it occurs. For example, moving a wooden piece across a board holds different meanings depending on whether it takes place during a game of chess, draughts (checkers), or in a scenario where children unfamiliar with chess spontaneously invent their own rules. Similarly, understanding the meaning of language or actions in everyday life depends on recognizing the broader linguistic gestalt—the structured arrangement of social practices, activities, and expectations—within which they occur.

Grammatical investigation requires us to examine these contextual patterns to clarify how meaning arises praxeologically within different linguistic gestalts—how words gain significance through their role in our practices and forms of life. ⁸ Put more succinctly, we “describe language games” (Wittgenstein, 1953/2009, Section 486). In doing so, we emphasise that meaning is praxeologically produced because, as Wittgenstein remarked, “[we] call the whole, consisting of language and the activities into which it is woven, a “language-game” (Wittgenstein, 1953/2009, Section 7).

Consider a familiar game, such as football (soccer). Kicking a ball around a pitch with the aim of scoring by putting it between the goalposts is what constitutes a game of football. The significance of actions in football—passing, scoring, drifting offside—depends entirely on their role within the game. For example, touching the ball with your hand in-play is a ‘handball’ and a foul, unless you are the designated goalkeeper. This relationship is reciprocal, exemplifying what Köhler called ‘dependent part qualities’: the players’ actions constitute the game of football, yet it is only within the game that these actions acquire their specific identity. The identity of each action—whether a kick is a pass, an attempt on goal, a scored goal, or a foul—depends on the game as a whole. These actions are what they are because they occur within the game; their identity is inseparable from the game itself. Just as actions in football derive their meaning from their relationship to the game of which they are a part, so too do linguistic expressions derive their significance from the linguistic gestalts within which they are embedded.

In short, Wittgenstein spoke of language games to highlight how language use follows certain rules, customs, or conventions—rules that are embedded in the very activities they constitute. Some areas of language use are akin to institutionalized games with codified rules, like Association Football, Rugby Union, and Baseball, where formal rules are explicitly defined and regulated. Other areas function more like informal games, such as ‘tag’ or ‘catch,’ where the rules are fluid, locally negotiated, and adapted as the game unfolds.

Building on our earlier discussion of perceptual gestalts and the relationship between constituents and wholes, we can see a similar dynamic in language games. The same behaviour can have different meanings—and therefore constitute a different action—depending on the game being played and the rules that constitute it. Similarly, we can recognize different language games at play: giving orders, asking questions, praying, presenting a proof, offering an excuse, stating a hypothesis, expressing anger, telling a joke, and so on. The same string of words might have different meaning (including, possibly no sense at all) in different language games.

Consider the utterance “Mo died because we didn’t know their blood type”. This could take on different meanings depending on the context:

1. A statement: A doctor explaining a case in a malpractice tribunal.

2. A question: A doctor asking a colleague whether Mo’s death was due to an unknown blood type, anticipating a possible alternative explanation (e.g., “No! We eventually worked out Mo’s blood type, but the injuries were too severe to save them.”).

3. A joke: “Mo died because we didn’t know their blood type. While breathing their last breath, Mo kept saying ‘be positive!’ but it’s hard without Mo.”

Wittgensteinian language games, therefore, are types of linguistic gestalts, where the linguistic gestalt or game (e.g., joking) is the whole, and the expressions that constitute it are the constituents. In examples 1–3, we see three distinct linguistic gestalts at work: stating facts, asking questions, and joking. To illustrate this point more directly with respect to psychology, let us return to our recurring example of grief and the funeral.

Grief can be understood as a form of emotional pain, distinct from the kind of pain caused by direct physical injury. For instance, the pain from striking one’s thumb with a hammer can, for certain purposes, be adequately explained in physiological terms: the hammerhead impacts the thumb, triggering nerve responses that travel to pain receptors in the brain. But what of the pain caused by hearing about the death of your good friend Mo? Here, in place of a material object, we have some words spoken to you by a friend. Although such pain is likely to be deeper and longer lasting than that of the hammer, an explanation from a physiological aspect is not going to be remotely adequate. How then, might we make sense of it?

The first step is to accept that ‘grief’ is not an abstract phenomenon with a particular meaning. Rather, its meaning is tied to those words having been spoken to you, the person with this relationship to Mo. If that same person died, but in a possible world in which you had not known them, their death would not bring about such pain. The grief is thus brought about, perhaps even constituted by, the meaning the words have for you. As such, it is an explicitly personal (not sub-personal) phenomenon. Second, the grief is constituted by the meaning it has as grief-in-response-to-this-event. The meaning of the event and the meaning of the response are internally related. In other words, the grief is always directed at something; it has intentionality. There is no grief module or container, so to speak, which is given content by events. It can be tempting to think that just as when we pour the tea out of a cup of tea we are left with a cup, when we ‘pour’ “for Mo” out of “your grief for Mo” we are left with a ‘grief container’ (or module). However, considering our gestalt-inspired Wittgensteinian remarks, the content-container analogy, which might be operative in the background as an unacknowledged assumption, is misleading. The grief only has sense as grief for someone (in this case Mo). In light of these points, therefore, we argue one should conceive of grief in terms of linguistic gestalts. In so doing, we do not fall into the trap of separating ‘grief’ from what it is directed at, nor do we misunderstand grief as a sub-personal process, actuated module, or structure. This respecification, we further argue, holds for all psychological phenomena.

Respecifying psychological phenomena in terms of linguistic gestalts means understanding them through the meaning of what is said and done within a praxeologically produced situation. In PA, the task is not to explain but to describe the relevant linguistic gestalt, thereby clarifying the phenomenon under investigation. PA is, therefore, not a strictly empirical form of investigation, but a kind of conceptual investigation that Wittgenstein called grammatical.

The conceptual foundations of PA – a praxeological development of Berlin School gestalt psychology into Wittgensteinian investigative ordinary language philosophy – lead to a respecification of psychological phenomena in terms of linguistic gestalts. Moreover, they provide the methodology with three central characteristics: (i) PA is situated (not abstracted), (ii) PA is descriptive (not explanatory), and (iii) PA is grammatical (not empirical). In the next section, we explicate these three characteristics in more detail and, in so doing, map PA’s relation to and differences from existing qualitative approaches, making a case for its distinctiveness.

The Central Characteristics of Praxeological Analysis

Situated Not Abstracted

The first characteristic of PA – that it is situated – means that it is aimed at investigating psychological phenomena in the contexts in which they manifest in everyday life. Returning to the phenomenon of grief, that could mean observing family members at a funeral, engaging in a longer project of investigation with a grieving person as they go about their everyday life, etc. It would not entail seeking retrospective accounts of experiences and perceptions of grieving through, for example, a set of interviews; or, even more abstractly, to subject participants who are grieving to an fMRI scan. The case for PA rests on the claim that psychological phenomena cannot be meaningfully studied in artificial contexts that alter the very behaviours they seek to investigate. Changing the setting in which an experience makes sense necessarily transforms both behavior and the psychological phenomenon itself (“Sit still and stop rubbing your head in your hands, or you’ll dislodge the sensors.” “I’m sorry, I know this is difficult for you, take as much time as you need to compose yourself before answering the question.” etc.).

The situated character of PA places it in relation to a number of extant qualitative methodologies, e.g., Ethnography, Grounded Theory (and its derivatives such as Situational Analysis), Ethnomethodology, Conversation Analysis, Discursive Psychology, and certain types of Action Research. Conversely, it sets it apart from methodologies that almost always (albeit not necessarily, if one’s methodological definition is excessively liberal) seek to investigate psychological phenomena through retrospective or reflective accounts: e.g., Interpretative Phenomenological Analysis (IPA), Descriptive Phenomenology, and most applications of the hyper-flexible Thematic Analysis (although, of course, Thematic Analysis needs additional conceptual underpinning to become a fully-fledged methodology). IPA presents an interesting contrast because, while it is largely non-situated, it claims a connection to the phenomenological tradition. However, its primary data collection method—semi-structured interviews—removes participants from the flow of absorbed coping, encouraging them to retrospectively reconstruct their experiences. This stands in tension with phenomenology’s emphasis on situated experience. This generates a tension between IPA’s de facto methods and its claims to being phenomenological. Some phenomenologists, such as Zahavi (2019), have gone so far as to argue that IPA, although clearly qualitative, non-reductive, and focussed on providing first-person descriptions, is not really phenomenological at all but amounts to something more like subjectivity analysis. From this perspective, a genuinely phenomenological approach must be grounded in key features: a critique of scientism, an appreciation of the lifeworld, an open-minded attitude, and, crucially, an investigation of the structures and conditions that make experience possible.

Descriptive Not Explanatory

The second characteristic of PA—its descriptive nature—follows from its respecification of psychological phenomena as linguistic gestalts. In a linguistic gestalt, meaning arises from the interplay between what we say and do and the situation in which those actions occur. Since meaning is embedded in context, there is no need for an additional explanatory framework beyond a careful description of how the phenomenon manifests in practice. All that is required for making sense of the phenomenon in question is to describe the particular instances in which it manifests. This highlights an important issue with psychological investigations: although in most instances we are able to make sense of local examples of a psychological phenomenon, global theoretical explanations can almost always be undercut. For example, although there are many competing and unreconcilable theories of conscious experience, as a rule, we have no problem making sense of conscious experience in everyday life. ⁹ Problems with discussing conscious experience only arise when we separate it from the contexts in which particular experiences make sense and try to explain it as an abstract, general phenomenon. The inadmissibility of further explanation in PA gives it a particular investigative orientation: unlike most (explanatory) qualitative approaches to psychology, which purportedly result in knowledge in the form of discovery, PA’s descriptive approach results in understanding in the form of realisation.

Knowledge in the form of discovery creates new knowledge that perhaps we could not even have conceived of, eliciting a response of the form “who would have thought of that!” (Hacker, 2013, p. 110). Although many qualitative psychologists work within this paradigm, this kind of response largely manifests successfully through explanatory theories in the natural sciences, which have both strong consensus and agreed upon criteria by which such consensus can be achieved in the first place; for example, the geological theory of plate tectonics which proposes the idea that the ground beneath our feet is not, as one might assume, stable.

In contrast to knowledge in the form of discovery, understanding in the form of realisation creates no new knowledge but instead brings us to a realisation about something we already knew, though perhaps implicitly in the form of know-how. These kinds of realisatons might be more likely to elicit a response of the form “Of course! I should have thought of that” (Hacker, 2013, p. 110). For example, in an investigation into the puzzle of what in addition to a bodily movement is involved in an action, the philosopher Frank Ebersole (2001c) carefully works through a range of everyday examples in order to bring about the realisation that, although we might be tempted to assume that actions are bodily movements plus intention, careful analysis of the situated use of the words ‘action’ and ‘movement’ reminds us (reminds because it is we who use these terms when we talk) that bodily movements and actions are just different ways of describing the same thing from different perspectives and for different purposes. Realisations, unlike discoveries, emerge through careful examination of specific instances and do not seek to produce general theories or explanations. Nevertheless, such realisations are valuable in clarifying and understanding a phenomenon, and in helping us to overcome misleading assumptions about the nature of the phenomenon.

This point is a philosophical one: what PA provides is a clarification of grammar by describing the use of language in situations. One might retort that such clarifications can be readily described as providing new information or knowledge. However, the reason for resisting this and saying instead, following Wittgenstein, that what grammatical investigation provides is not new knowledge is that we already implicitly know the rules of grammar because we are competent users of the language. What a grammatical investigation does is make the implicit knowledge explicit and, hopefully clear (or perspicuous) for those who might benefit from this clarity. Therefore, it is, in this sense, misleading to depict the ‘products’ of a grammatical investigation as new knowledge because the knowledge is there all along embedded in the practices of members of society. A grammatical investigation seeks only to make this practical know-how clear and surveyable for the specific purposes of understanding psychological phenomena and dissolving philosophical confusion.

The descriptive character of PA sets it apart from some methodologies that its first characteristic – situatedness – aligned it with. For example, the range of methodologies that explicitly aim to generate explanation from the ‘data’ (in some instances through the development of a theory), such as Grounded Theory, Situational Analysis, and some versions of Action Research. It also differentiates PA from approaches that rely on pre-existing theoretical frameworks for explanation, such as some phenomenological methods, Foucauldian Discourse Analysis, and Narrative Analysis. Unlike these approaches, PA remains strictly descriptive, focusing on how meaning is enacted in practice rather than theorizing about it.

Grammatical Not Empirical

The third characteristic of PA—its grammatical nature—is, like its descriptive character, rooted in Wittgensteinian philosophy. It is this final characteristic that affords PA its distinctiveness in the methodological landscape. Wittgenstein used the term ‘grammatical’ to refer not just to the structural relationships of words in a language, but the wider rules, customs, and conventions that have to be settled for an expression (words, actions, etc.,) to make sense in everyday life. This wider notion of grammar can only be made sense of through careful reflection on detailed descriptions of everyday situations (and their linguistic gestalts). Importantly, the sense one makes of such descriptions (i.e., the realisations arrived at through grammatical investigation) is a form of conceptual clarification, or what Wittgenstein called a “perspicuous presentation” (1953/1958, Section 122). ¹⁰ This mode of investigation does not rely on empirical criteria but merely on clarifying the appropriateness of the linguistic gestalt in question. So, for example, thinking back to the duck-rabbit picture, no amount of pointing to the lines on the page seen as elements will help to clarify the picture – they have to be conceptually clarified as rabbit ears (rather than a duck bill) to make sense as part of a rabbit. In another, more everyday, context, confusing a teenage goth for a grieving sibling is merely an elementary mistake, which cannot be overcome by more observation of teenage goths or grieving siblings abstracted from the context in which they are encountered. The distinctive character of grammatical investigation is captured by Wittgenstein’s (1953/2009) remark that “[philosophical] problems are solved, not by coming up with new discoveries, but by assembling what we have long been familiar with” (PI §109), and that such a mode of investigation “leaves everything as it is” (PI §124). In other words, you are already implicitly familiar with grammatical distinctions because you regularly make them in the flow of everyday life. What grammatical investigation does is remind you of these distinctions by making them explicit and presenting them perspicuously.

The grammatical (i.e., conceptual not empirical) character of PA has a significant methodological consequence, and it is this that identifies PA as a distinctive qualitative methodology. Because descriptions in PA serve only the purpose of conceptual clarification—providing perspicuous presentations of psychological phenomena—they do not need to be descriptions of actual situations. While actual cases remain integral to PA, imagined or fictionalized scenarios can also serve as valuable tools for revealing conceptual distinctions. Furthermore, as outlined in the next section, descriptions of imagined or fictionalised situations are often very useful in conceptual clarification, as long as they can be reasonably conceived as realistic. In a previous paper (2022) we explicated a narrower form of grammatical investigation for philosophy – termed Investigative Ordinary Language Philosophy – that relies solely on imagined and fictionalised descriptions. PA represents a more complete explication of grammatical investigation, where the imagined and fictionalised situations are more likely to take the role of “intermediate cases” (Wittgenstein, 1953/1958, Section 122) ¹¹ alongside the use of actual descriptions and, as such, situates PA as a qualitative method of analysis, to be used wherever qualitative research is undertaken, and not purely a method of philosophical analysis.

The grammatical character of PA sets it apart from existing methodologies that arguably share its first two characteristics: Ethnography, Ethnomethodology, Conversation Analysis, and Discursive Psychology. In the methodological literature there is really no live debate on whether ethnography uses descriptions of non-actual situations. We thus focus here on the latter approaches.

Ethnomethodology (EM) is interested in how members co-produce social facts and social structures in interaction and, in the process, make their production of those social phenomena witnessable; or in Garfinkel’s terms, accountable. Analogously with PA, there is a synergistic relation, or reciprocity, between our actions which identify us as members of social units and the identity of the social unit as a whole. Conversation Analysis (CA) emerges from these key features of EM, but with the focus on conversation as the social phenomenon par excellence. PA is heavily indebted to our Wittgensteinian ‘misreading’ of foundational work in EMCA, including Garfinkel (1967) and Sacks (1995). However, despite the striking similarity of some of the analysis found in Sacks’ early lectures, and marginal accounts of Ethnomethodology (such as the work of the Manchester school of EMCA) ¹² to grammatical investigation, EMCA still trades exclusively in descriptions of actual situations.

Discursive Psychology (DP) emerged in the late 1980s, and while there are a variety of approaches that have taken the name (e.g., Harré & Stearns, 1995; Potter & Edwards, 1992), DP has more recently developed into a version of Conversation Analysis which focusses specifically on “practices that use mental terms, and how those words are operating in those practices” (Potter & Edwards, 2003, p. 173). Potter (occasionally with Edwards) has written extensively on DP’s debt to ordinary language philosophy, and specifically the work of Wittgenstein and Austin (e.g., Potter, 2001), in addition to its EMCA heritage (Potter & Edwards, 2013). However, although there are obvious overlapping features, PA differs from DP primarily because although there is some debate over whether it is correct to say that DP is empirical, Potter and Edwards explicitly emphasise the importance of analysing samples of actual discourse (Potter & Edwards, 2003). Thus, because both EMCA and DP exclude non-actual situations, PA stands alone as the only fully developed approach to grammatical investigation in the human and social sciences—one that uniquely integrates conceptual clarification with qualitative methodology.

What Praxeological Analysis Consists of

Having now worked through the three central characteristics of PA, we finish this section by concisely stating what PA consists of: PA is a grammatical form of investigation that aims to understand phenomena through bringing about realisations derived from descriptions of actual, imagined, and fictionalised situations. This concise account of PA guides the direction of and methodological choices in a PA study.

Doing Praxeological Analysis

In this section we outline how to conduct an PA study through explicating its six stages.

The Six Stages of Praxeological Analysis

1) Respecify the psychological phenomenon as situated and praxeologically produced.

2) Produce descriptions of actual situations in which the phenomenon manifests.

3) Reflect on the descriptions of actual situations and what they tell us about the situated meaning of the phenomenon.

4) Produce descriptions of imagined and fictionalised variations on the actual situations as intermediate cases.

5) Reflect on the descriptions of imagined and fictionalised intermediate cases to gain a deeper understanding.

6) Comment on the study findings with respect to existing accounts.

Box 1: The six stages of PA

While we do not have the space here to report an entire investigation, we shall explicate the six stages of PA through focus on aspects of an investigation into pain talk. We start with a ubiquitous picture of feeling pain as a mental reaching across the body and pain as an object to be felt. This way of beginning with pervasive pictures of the way things are is a characteristic starting point for PA, insofar as such pictures encapsulate a way of thinking that seems appealing (even natural or inevitable) but which might lead us to say something we do not want to say or serve to constrain our thinking; for example, that if feeling pain involves reaching across the body, I must be someplace else than the pain.

Stage 1: Respecify the Psychological Phenomenon as Situated and Praxeologically Produced

Although seemingly simple, the first stage of PA is crucial because it ensures that the investigation is oriented toward the correct unit of analysis: linguistic gestalts. In PA, psychological phenomena are not treated as internal cognitive states or mechanistic processes but are instead respecified as linguistic gestalts—patterns of meaning that manifest in everyday life through interaction. An important aspect of this respecification is that PA does not start with reference to existing psychological theory or definitions of key terms. In support of this respecification, researchers should, in line with many other qualitative approaches, aim to isolate their investigation from existing psychological doctrine by adopting an open attitude. This is a conscious practical move in order to hold back the natural desire to theorise and explain. As the investigative ordinary language philosopher Frank Ebersole (2002, p. 325) notes, this is important because “[these] theories that rush in not only make me distort the facts, they make me blind to the very facts they have led me to distort.” This does not mean that researchers must forget their theoretical knowledge of the phenomenon. Rather, they should temporarily set it aside, allowing the descriptive process to guide the investigation without theoretical preconceptions shaping the grammatical elucidations. Such an approach is related to the notion of ‘bracketing’ and similar terms found in other methodologies.

Stage 2: Produce Descriptions of Actual Situations in Which the Phenomenon Manifests

The second stage of PA entails producing descriptions of actual situations in which the phenomenon in question manifests. While grammatical investigation does not inherently require actual descriptions, PA begins with them because they offer a concrete foundation for analysis—particularly for phenomena that researchers may have little prior experience with. Starting with real-world descriptions ensures that the investigation remains grounded in the ways language is actually used before considering broader conceptual clarifications. Descriptions of actual situations can be produced in two primary ways: (1) through direct participant observation, or (2) by collecting new or existing audio, video, or written records of interaction. There are extensive existing explanations of these approaches, which provide useful practical advice (see DeWalt & DeWalt, 2002; Hammersley & Atkinson, 2007; Kiyimba et al., 2019; Ten Have, 2007; Wolcott, 2005). Given PA’s foundation in linguistic gestalts, descriptions should focus on small-scale, interactional settings, paying particular attention to verbatim speech, nonverbal expressions, and gestures. ¹³ To illustrate how descriptions in PA might be structured, we present two examples from doctor-patient interactions related to the experience of pain. These examples highlight how language, interaction, and context shape the meaning of pain in clinical encounters. ¹⁴

Example 1

Example 1 features Dr. W, an experienced General Practitioner who has worked at the same practice for over 20 years. The patient (P¹) has been seeing Dr. W for several months regarding their symptoms, and they have an established doctor-patient relationship.

Example 2

Example 2 features Dr. M, another experienced General Practitioner with a longstanding relationship with their patient (P²). Over the years, they have developed a strong rapport, which shapes their clinical interactions.

Example 1

Dr W. A month gone, maybe more, what’s happened?

P¹. I don’t know. it’s not right.

Dr W. You were finding it hard to do stuff?

[They talk a bit more. Dr W. notes how he thinks P¹ is a “do-er”]

P¹. I had a shooting pain in my neck… going from my groin to my hip, down my leg… I know it’s not completely right… I just go to work… blood tests ok?

Dr. W. Yes, we’ll get to those… Why I said you are a do-er is that you just get on with it.

[They talk a bit more about P¹’s symptoms]

P¹. … That’s just annoying me.

Dr. W. Pain?

P¹. I get the pain in the groin area… not as bad as before… Is that sciatica do you think?

Example 2

[Dr M. and P² move on to talk about diabetes]

P². I’m not borderline diabetes?

Dr M. No, I don’t think so… let me double-check that… no… glucose was fine in June.

P². Sometime back, I did have pain in my chest… Dr M was concerned… Angina… I’m not am I?

Dr M. No, that test Professor B did was really helpful… [your] breathlessness is mostly to do with fitness.

Stage 3: Reflect on the Descriptions of Actual Situations and What They Tell Us About the Situated Meaning of the Phenomenon

The third stage of PA requires little procedural instruction but is, nonetheless, an important stage in the methodology. After compiling descriptions of actual situations, researchers should engage in careful reading and reflection, attending to linguistic patterns, implicit assumptions, and unexpected aspects of interaction. The goal is to allow meaningful realizations to emerge organically, rather than imposing predefined analytic categories. In line with approaches such as EMCA and DP, particular focus should be given to the words used in interaction and on the seemingly mundane aspects of the situations that we might ordinarily overlook. Unlike many other qualitative methodologies, this process does not require or entail a formal analysis or coding technique. Instead, researchers should pay close attention to the descriptions non-formally.

The realisations that are brought about in Stage 3 are not likely to bring any definite conclusion to the investigation. Rather, they will be intermediate realisations, which will bring about further questions and puzzles to be considered in Stage 4. For instance, considering our investigation into “feeling pain”, a striking realization might emerge: despite the focus on “feeling,” participants in both examples do not describe themselves as “feeling” pain but rather as “having” pain. This shift in language—e.g., “I had a shooting pain in my neck,” “I get the pain in the groin area,” “I did have pain in my chest”—suggests that the way pain is experienced and described is not necessarily aligned with the original, perhaps theory-informed, framing of the investigation.

Stage 4: Produce Descriptions of Imagined and Fictional Variations on the Situation as Intermediate Cases

As noted in Stage 3, intermediate realizations rarely lead to definitive conclusions but instead generate further questions and conceptual puzzles. For instance, in the investigation into pain, a key puzzle might arise: Why do people typically say they “have” pain rather than “feel” pain in everyday contexts? The fourth stage of PA entails producing descriptions of imagined and fictional – albeit realistic – situations that will hopefully help to clarify the questions and puzzles that emerged from the intermediate realisations in Stage 3. These descriptions are termed intermediate cases and are produced in the service of our grammatical investigation, so that we might explore different use-cases. Imagined and fictionalized descriptions are particularly valuable for clarification because, unlike actual cases, they can be deliberately selected or constructed to address specific conceptual puzzles that emerge in the investigation. To illustrate how imagined cases function in PA, consider the following three examples related to the investigation into pain. These cases explore different linguistic contexts in which the concepts of “having” and “feeling” pain might arise. ¹⁵

Example 3

While hiking with a friend, I limp to a halt and sit on a log holding my knee. “What’s wrong?” “I have a terrible pain in the knee.” (I do not say, “I feel a terrible pain in the knee.”)

Example 4

A man has suffered paralysis of the legs and is slowly recovering. Every day the doctor touches, probes, moves his legs. He asks, “Do you feel anything?” One day the patient says, “Yes. I feel a deep pain in the ankle.”

Example 5

I have injured a leg and am suffering from an unbearable pain in the knee. I am given a local anaesthetic and gradually the pain subsides. My leg becomes completely numb. Later the doctor asks “How is it now?” “I can feel the pain again.” Or “I can feel that pain again” My sensitivity is returning to normal.

Stage 5: Reflect on the Descriptions of Imagined and Fictionalised Intermediate Cases to Gain a Deeper Understanding

After producing descriptions of imagined and fictional/fictionalised situations (i.e., intermediate cases) in Stage 4, Stage 5 involves reflecting on those descriptions, particularly with respect to the questions and puzzles that emerged in Stage 3. As in Stage 3, this process does not involve formal analysis or coding techniques. Instead, researchers should focus on identifying key similarities and differences across descriptions, paying attention to subtle shifts in meaning and usage. For example, in the investigation into feeling pain, the realisation may emerge that ‘feeling’ pain is only related to very particular instances involving sensibility or numbness. This suggests that “feeling” pain and “having” pain are not interchangeable. As Ebersole (2001a) notes, “the background for ‘I feel a pain’ must be carefully prepared,” indicating that the use of “feeling” in relation to pain is contextually constrained. The misleading picture we began with—of feeling pain as a mental act of reaching across the body—loses its grip once we recognize that “feeling pain” only makes sense in specific contexts related to sensibility and numbness. In most cases, therefore, it makes no sense to say that we feel pain; pain is simply something we have or don’t have. Moreover, situations in which we do say “I feel the pain” are shown not to be equivalent to “I feel the pen (in my pocket)” or “I feel the egg (in my hand)” but are rather more akin to someone reporting that they are recovering their sense of/ability to touch/smell/taste. Given that ‘having’ pain is an explicitly personal (not sub-personal) capacity, the problem of locating pain in specific bodily locations and then transmitting it to an I thought to be located somewhere else (which we encountered when conceiving pain as being felt) dissolves as does the propensity to think of pain as a thing which the word ‘pain’ names. By dissolving this thought-constraining picture of “pain,” we avoid the conceptual traps that, if left unexamined, could misdirect research and obscure the relationship between the self and its experience of pain.

Stage 6: Comment on the Study Findings With Respect to Existing Accounts

Although Stage 5 concludes the grammatical investigation central to PA, Stage 6 reintroduces the final realizations into the broader landscape of existing theoretical accounts—accounts that were intentionally bracketed during the investigation. This may include reflecting on existing theories or frameworks in light of the new realisations. It may include considerations of conceptual and practical implications.

The distinction between ‘having’ and ‘feeling’ pain reveals conceptual tensions in contemporary psychology and healthcare. One such tension arises in dominant neuroscientific theories of pain, such as Melzack’s hugely influential neuromatrix theory (1999), which conceptualises pain as a sensory-perceptual phenomenon transmitted through neurological pathways and then subject to modulation by cognitive and emotional factors. Such models rely on an implicit picture of pain as something that is ‘felt’ in the same way that we might ‘feel’ an object’s texture or temperature. However, our grammatical investigation suggests that ‘feeling’ pain is only used in particular cases where there is some question about sensibility, such as when numbness is subsiding. In most cases, people speak of pain as something they ‘have’ rather than ‘feel’. This suggests that framing pain as a kind of perceptual object perceived by sub-personal networks may overemphasise the sensory aspect while underplaying the embedded, situated character of pain in everyday life.

Imagine you’re in pain. You say “I have a pain”, but not “I feel a pain” unless you’re being tested for numbness. This suggests that pain is not just something we detect like a thermometer detects temperature, or a fire alarm detects smoke—it’s woven into how we talk about experience. Melzack’s theory proposes that pain is generated by a neural matrix in the brain, producing a unique “neurosignature”—a distinct pattern of neural activity that defines pain. In this view, pain functions like a signal the body generates, much like a computer executing a pre-programmed response to a stimulus.

Our Ebersole-inspired praxeological analysis leads us to respond that pain is not a signal. Pain is not a “thing” inside your brain. Rather, pain talk is a way we express and respond to bodily experiences—it is not something that can be mapped like a fixed transmission between sender and receiver. Melzack’s theory assumes that pain is a biological function of subpersonal systems, but we might respond that reducing it to signals inside the nervous system comes at a cost: it strips pain of its embeddedness in human life and social interactions, it strips it of the meaning it has in our lives.

So, while Melzack is trying to scientifically “explain” pain as a neurological process, our study would suggest that this is to look in the wrong place and, in doing so, miss important aspects of the phenomenon of pain. ‘Pain’ isn’t the name for something we locate inside us—it’s part of how we talk, express, and live with bodily experiences in human interactions.

Stage 6, although not integral to the grammatical character of PA, is nonetheless an important stage because situating the findings from a PA study back into the wider landscape helps to communicate the relevance and importance of its findings. The foregoing reflections demonstrate how the investigation of ‘feeling pain’ through PA challenges a number of entrenched assumptions in psychology, inviting researchers to reconsider the conceptual foundations of their work and to align their methods more closely with the lived realities of those they study.

A summary of the Investigative Process in Praxeological Analysis

1) Respecify the relevant psychological phenomenon as situated and praxeologically produced.

2) Produce descriptions of actual situations in which the psychological phenomenon manifests, using either participant observation or by collating new or existing audio, video or written accounts.

3) Reflect on what the descriptions produced in (2) mean for understanding the psychological phenomenon, in order to bring about intermediate realisations – this will likely bring about further questions and puzzles to be considered in (4) and (5).

4) Produce descriptions of imagined and fictionalised situations (intermediate cases) that help to clarify the questions and puzzles that emerged from the provisional realisations brought about in (3).

5) Reflect on the descriptions produced in (4), particularly with respect to the questions and puzzles that emerged in (3). This will bring about the study’s final realisations.

6) Comment on the study findings with respect to existing accounts, in order to situate the findings in the wider landscape.

Box 2: A summary of the investigative process in PA

Conclusion

This paper has introduced Praxeological Analysis (PA) as a new qualitative methodology in psychology. While our focus has been on psychology—reflecting both authors’ current disciplinary affiliations—PA is not limited to this field. Drawing on our interdisciplinary expertise in philosophy, health, sociology, and political studies, we argue that PA has broad applicability across the social, human, health, and education sciences.

In this introductory article, we have argued that once psychological phenomena are respecified in terms of linguistic gestalts, the appropriate mode of investigation is what Wittgenstein termed grammatical. This mode of investigation entails bringing about realisations derived from descriptions of actual, imagined and fictionalised situations. In this concluding section, we discuss how PA – as a complete mode of grammatical investigation – can benefit psychology as a discipline.

First, it is important to note that psychologists will only see benefit in PA if they are open to and accept the Wittgensteinian respecification of psychological phenomena on which it is grounded. As can be inferred from our explication of PA, there are broadly two key aspects involved in such a respecification: first, that psychological phenomena are inextricably entangled with meaning; and second, that such meaning is situated in the linguistic gestalts through which we live our everyday lives. Therefore, abstraction from the situations of our everyday life untenably alters the psychological phenomenon under investigation. It is a rare qualitative psychologist who does not accept the first premise. Moreover, many qualitative psychologists already accept the idea that psychological phenomena are situated, even if they do not explicitly conceptualize them in terms of linguistic gestalts. This suggests that PA’s respecification aligns with perspectives that are already widely accepted, making it a natural extension rather than a radical departure. Indeed, more specifically, there a number of extant conceptions of psychology that either explicitly or implicitly accord with our Wittgensteinian account. ¹⁶

With many qualitative psychologists likely being open to our respecification of psychological phenomena, the real benefit of PA comes from convincing them that it provides the most appropriate investigative methodological approach. We finish by noting that PA does this by bringing together two different but related Wittgensteinian approaches to psychological investigation. First, it employs a grammatical mode of investigation that relies on descriptions of actual situations, as seen in certain interpretations of EMCA and DP. Second, it incorporates Wittgenstein’s method of conceptual clarification, which uses descriptions of imagined and fictionalized situations—an approach that has remained on the margins of the philosophy of psychology (e.g., Racine & Slaney, 2013). By integrating practical empirical skills with a robust grammatical and conceptual framework, PA offers a novel and compelling methodological approach—one that has the potential to reshape psychological research and extend its conceptual foundations.