In an article currently available online first at American Journal of Cultural Sociology, Christina Simko and Jeff Olick (hereafter S&O) propose and develop a new dimensional characterization of cultural phenomena, what they refer to as a “four facet” model of culture. On the one hand, they distinguish between cultural phenomena along a dimension separating (public) discourses and (action-oriented) practices. These are cross-cut by a second dimension, differentiating between “implicit” and “explicit” culture, yielding explicit versus explicit discourses and implicit versus explicit practices. Their “dimensionalizing” approach aims to provide a summary framework for the classification of cultural kinds, and as such, dovetails with emerging discussions on the ontology of culture that have been brought up in this blog (see, e.g., here, here, here, and here) and elsewhere (e.g., Rotolo 2020).
In this post, I consider the question of how S&O’s proposed typology links up to other dimensional distinctions previously discussed. In particular, I focus on the “implicit/explicit” property dimension. This is a potentially productive analytic endeavor because even though the idea of “implicit culture” is probably as old as the anthropological culture concept itself (Bidney 1968), it continues to be hampered by analytical fuzziness and inconsistency in application. In what follows, I ask: What sort of ontological claim do we make when we say of a cultural kind that it is implicit or explicit? Does it matter what other properties the kind in question possesses? In other words, is calling culture “implicit” or “explicit” the same type of predication when we talk about personal and public cultural kinds?
To anticipate, my argument is that “implicitness” claims about culture are of a different nature as other types of claims (e.g., locational or compositional). Instead, I argue that implicitness is a relational (and thus observer relative) property and thus presupposes cognitive agents with an awareness (or unawareness) relation vis a vis the given cultural kind. This makes the use of the implicitness dimension as a way to characterize cultural kinds a trickier affair than when using other dimensions, as is deciding whether the notion of “implicit discursive (public) culture” is ontologically coherent. I conclude that when it comes to characterizing cultural-cognitive kinds (learned and internalized by people), the implicitness/explicitness points to a core property of the kind. In contrast, when characterizing external (non-cognitive) cultural kinds, they get at incidental, non-intrinsic properties.
This means that personal culture is implicit in a way different from public culture. Analysts get into trouble when they extend the version of implicitness appropriate for cultural-cognitive (mentally represented) cultural kinds to public (non-mental) culture (Turner 2014). Instead, what we need to do is come up with a version of “implicit” that makes sense for public cultural kinds, without stirring up the ghosts of collective minds and related problematic assumptions.
Simko and Olick’s Four Facet Model
As already noted, S&O’s four facet model distinguishes between culture as practice versus culture as discourse and then proposes implicit and explicit variants of each. The first (discourse/practice) distinction implies that culture is either located on the side of public meaning and signification (discourse) or on the side of (inter)personal action (practice). This is now a fairly standard differentiation in cultural analysis, going back to the usual langue/parole distinction in Saussurean semiotics (Sewell 2005), and appearing in related forms in a variety of recent “post-structuralist” proposals (e.g., Biernacki 2000; Swidler 2001; Sewell 2005; Turner 2007).
From the perspective of a more differentiated cultural ontology, might complain that the practice/discourse distinction is too coarse, combining locational and compositional claims. We could also point out that there are additional cultural kinds at the personal level beyond practices (e.g., declarative personal culture) and other public cultural kinds beyond discourses (Rinaldo and Guhin 2019). Nevertheless, the broad distinction along the discourse/practice dimension makes overall sense, at least in terms of partitioning between what are consensually regarded as distinct cultural kinds.
It is in deploying the second (implicit/explicit) distinction that things get more complicated. One problem is that satisfactory definitions of “implicit culture” are hard to find. The ones on offer tend to err on the side of over-inclusiveness, making them analytically unwieldy. For instance, in a programmatic review written during the rise of cultural analysis in American sociology, Wuthnow and Witten (1988: 50) defined implicit culture as:
…[C]ulture [that] appears to be “built into” all social relations, constituting the underlying assumptions and expectations on which social interaction depends. One thinks of Polanyi’s concept of tacit knowledge as an example, or Parsons’s emphasis on the normative underpinnings of social action, and more generally the idea of axiological principles that govern civilizations or the notions that derive from Weberian sociology about the role of beliefs and presuppositions as guides for social behavior. In these views, culture tends to be regarded not as an explicit product but as a prefiguration or ground of social relations.
This version of implicit is hard to distinguish from the package of distinctions and properties of cultural kinds that influential lines of theory usually sweep under the notion of the “tacit.” As Stephen Turner (2014: 67) has argued, these are typically referred to in terms of, among other things, “Paradigms, Weltanschauungen, Presuppositions, Structures of Consciousness or Meaning, Collective Consciousness, Systems of Collective Representations, Tacit Knowledge.” When these are thought of as both “shared” and distinctive of a given social group or institutional sphere, we get the (problematic) notion of the “collective tacit”; it is clear (in the allusions to Polanyi and talk of “underlying assumptions”) that Wuthnow and Witten are thinking of implicit culture in this way. This is a (presumed) cultural kind that has carried a tremendous explanatory burden across distinct lines of cultural and social theory; this a load so large, according to Turner (2014), that it cannot be coherently met without begging the question and introducing a myriad of analytic inconsistencies and irremediable problems (see also Turner 1994).
S&O recognize their analytic debt to this line of work in developing their own version of the implicit/explicit culture distinction, noting that Wuthnow and Witten had already pointed to this dimension as a major one in cultural theory. However, in contrast to W&W who, as we have seen, conceptualized this in terms of a “collective tacit” model, S&O link the explicit/implicit distinction to recent work in cultural analysis (mine) done from a more cognitive perspective:
To take a more contemporary example, Lizardo and Strand (2010) make the implicit/explicit distinction a cornerstone of their effort to relate toolkit theory to cognitive approaches, or what they call “strong practice theory,” and, as already discussed, Lizardo (2017) extends their work in his distinction between declarative and nondeclarative forms of personal culture. More often than not, however, and as Lizardo (2017) also argues, the implicit/explicit distinction remains unarticulated and unnoticed—again, generating conflict between (what appears to be) competing conceptions of culture.
One question that arises in the wake of this conceptual linkage is whether the implicit/explicit distinction that gets mapped into cultural cognitive differentiation between declarative and nondeclarative culture when the ontological claim is that this culture is located in people is comparable to the same claim being made of a “collective tacit” cultural kind. In a previous post on cultural practices, I argued that the two distinctions are not the same, and that “collective tacit” kinds face an ontological problem of unclear location, making them a possible candidate for elimination.
One consideration is that, from a cultural-cognitive point of view, to say a cultural kind is implicit is saying something about the “format” in which that cultural kind is embodied in the human agent; not about its individual or collective status (which is incidental). That is, as either a declarative (consciously accessible and linguistically expressible) representations or as nondeclarative (hard or impossible to articulate) skills or dispositions. As such implicit/explicit is not a general property applicable to all cultural kinds, but a restricted property only meaningfully applicable to cultural-cognitive kinds, since these are the ones that can be learned and internalized by people (what I called personal culture in Lizardo 2017).
This means that the declarative/nondeclarative distinction (for personal culture) can be remapped in S&O’s four-facet model as a distinction between practical-explicit and practical-implicit without (too much) analytic loss. In fact, their examples of practical implicit cultural kinds” taken-for-granted assumptions that intuitively guide action and interaction,” and practical explicit ones “accounts and justifications…explicitly stated rules, norms, and procedures” clearly show that they are getting at the same differentiation as I did in Lizardo (2017) here.
The issue is whether S&O’s “implicit-discursive” culture ends up being yet another version of postulating the existence of hypothetical “collective tacit” kinds. Unfortunately, it seems like that is precisely what it is, as their proposed examples reproduce the standard list of usual suspects isolated by critics of the notion such as Turner. Thus, implicit-discursive culture, according to S&O, includes:
…group- or societal-level cultural codes that can be readily mobilized to construct ideas. Implicit-discursive culture is the so-called “deep structure” from structuralism, Durkheim’s… “collective conscience,” langue as opposed to parole, Foucault’s conception of the discursive formation…or the episteme…and the binary codes…and generic templates…underlined in the strong program.
S&O go on to theorize each of the possible undirected types of relationships that we can observe between these four cultural elements (which turn out to be six ([4*(4-1)]/2).
In the remainder, I focus on a more targeted issue. Is “implicit/explicit” when used with reference to cultural-cognitive kinds internalized by people the same as “implicit/explicit” when applied to putative non-cognitive (or at least non-personal) cultural kinds such as discourses?
Let’s take a paradigmatic cultural-cognitive kind, such an internalized “schema.” Calling this cultural entity “implicit” is just to say that a lot of the structure of a schema, its component parts (e.g., slots, default fillers, structural constraints), its relational organization, and even the way we go about putting it to use to deal with real-world tasks (e.g., categorization) is not consciously available for people to report. That is, most schemas work in a nondeclarative way as a type of procedural memory (Rumelhart 1980). Only with great effort can one take even the simplest of schemas (e.g., for the concept of “buying” something) and “redescribe” into an explicit representation (D’Andrade 1991). This entails “duplicating” (a version of) the schema into this new explicit representation; when we go back to using the “buy” schema in everyday cognition, we fall back on the nondeclarative, procedurally represented one (Rumelhart 1980).
Accordingly, when we apply the predicate “implicit” to a cultural-cognitive kind we are making a statement as to how it is organized and represented in memory, and thus how it ends up being “used” in action (Lizardo 2017). A fundamental message of recent (post-practice theory) work in culture and cognition studies, is that a lot of the cultural cognitive entities that previously were thought as being represented and working exclusively via explicit pathways are also represented implicitly, and may do most of their work in directing action in this way too (Lizardo and Strand 2010).
But what does it mean to say like “group- or societal-level cultural codes” are implicit? Note that we cannot apply the “representational format” criterion off the shelf. The reason for this is that, since there is no “collective person” (or in older parlance, “collective mind”) that somehow internalizes and represents collective tacit kinds, then implicitness here cannot be an issue of mental representation, as there is no mind to do the representing or the learning. Note that if we just say, in search of ontological sanity, that the collective mind is just the set of all individual minds, then the “implicit discursive” just collapses into the “implicit practical” as we would only be talking about knowledge, schemas, know-how, or personal representations of collective codes internalized by (a lot of) people. Here, the “collective” aspect simply collapses into the weaker predication of “shared.”
But replication or sharing, as I argued before, is not the sort of cultural property that generates distinctions of kind and in fact, despite the sociological obsession with “sharing”; it may be one of the least analytically interesting (but not necessarily substantively inconsequential) of the cultural properties. A belief held by one person and a belief held by two people is still the same type of cultural-cognitive entity. The same goes for a skill; a skill known to a single animal is the same type of cultural entity when taught to a second animal, no magical transubstantiation into something else happens when a cultural-cognitive entity replicates in a population via learning, imitation, independent rediscovery, or other diffusion mechanism. So if “discursive” simply means shared discourses, and discourses are neuro-cognitive representations internalized by people, then shared or distributed discourses are not a cultural kind ontologically distinct from personal declarative discourses (Sperber 1996).
Note that the Wittgensteinian point about the so-called impossibility of a “private language” does not save the day here. First, the whole notion of impossibility is overblown here. A private language is entirely conceivable if, by that, we mean a mapping between lexical elements and cognitive meanings, such that the production of the first systematically evokes structured versions of the second (in the one person). In this sense, private languages can definitely exist, it’s just that for communicative and coordination purposes, they are pretty useless. But the more significant point is that in terms of what type of cultural cognitive entity it is, a language that goes from private to shared (e.g., a psychiatrist who “decodes” a schizophrenic idiolect) does not go from being one kind of thing to another by virtue of this sharing, although it does gain additional causal powers (e.g., affording communication) it didn’t have before.
Another issue, already intimated in bringing up the bugaboo of the “collective mind,” is that the property “implicit” is inherently epistemic, as it necessarily makes reference to a putative knower for whom the cultural cognitive kind is either implicit or explicit. In other words, we surmise a given cultural-cognitive kind is either implicit or explicit because of the way people go about becoming acquainted with it (e.g., knowing it). This observer-relativity of the property of “implicitness” makes it a non-starter when we want to use to refer cultural externalized in the material world in the form explicit linguistic markings, artifacts, codes, classification systems and so on, if this analogy is made directly to the same property ascribed to cultural-cognitive kinds.
Saving the Public Implicit
However, this does not mean that we should abandon the intuition that there is an analog to the “mental implicit” property when it comes to non-mental (externalized, public) cultural kinds. Consider a mundane complex artifact such as a watch (of the old-timey, analog variety). When exposed to such an object, we know that it has some explicit aspects (such as two-hands and the hour and minute markings); but we also know that it has a lot of aspects that remain implicit; for instance a complex system of gears and internal mechanisms that allow it to “work.” This implicit aspect of the watch could be made explicit by, for instance, breaking up the watch and reverse-engineering it.
Or consider that when, during the song “Miracles,” members of the hip hop duo Insane Clow Posse exclaim “fucking magnets, how they work?” they refer to aspects of the magnet’s functioning that remain forever implicit, due to their inability to conceive of how Maxwell’s electromagnetic theory provides an explicit account of their operation.
I propose this version of implicit is both analytically salvageable and ontologically non-problematic for public cultural kinds.
Note that there are two facets of this notion of implicitness. First, it refers to an important property of public cultural kinds but not to a “core” property. Second, just like with cultural-cognitive kinds, it is an observer-relative (relational) property; a public cultural or natural kind has implicit aspects only in relation to a given knower. Thus, magnets remain forever implicit for members of ICP, but not for trained physicists. Note that this observer-relativity of implicitness transfers to cultural kinds; a public cultural kind such as “country music” has many more implicit aspects for a casual fan of the genre than it does for a cultural sociologist trained in the intricacies of production of culture theory.
As such, a lot of public cultural kinds are like the artifactual kind “watch” or the natural kind “magnet,” having both explicit and implicit aspects, with the latter capable of being transformed into the former via some (interpretative, computational) procedure. For instance, consider a “text corpus.” This usually has a lot of explicit facets (e.g., we can read example documents from the text and get a sense of what it says), but it also has a lot of implicit aspects. For instance, this could be the fact that the word “obese” (and semantically related cognates) co-occurs, within a specific window of text, with a lot of words with negative valence to the extent that surpasses what we would expect by chance (Arseniev-Koehler & Foster 2020). These implicit elements of the text corpus can be extracted (and made explicit for an analyst) using a computational technique like neural network word embeddings. We might even use the process via which an algorithm “learns” the culture implicit in a text corpus as a model for how a human learner would extract implicit knowledge of concept-valence associations internalized as nondeclarative personal culture (Arseniev-Koehler & Foster 2020).
So “implicitness” in public culture is a different kind of property than implicitness in personal culture. Both have to do with epistemic relations between a knower and a given cultural kind, but the link between property and object is different. If somebody has an implicit bias (they associate obesity with laziness at an unconscious level), the implicitness part of the bias is not some incidental aspect of it; instead, it is essential to the cultural-cognitive kinds implicit attitudes are (Brownstein 2018). Implicitness is a core property of the whole attitude, pivotal to the way that it is learned (via multiple repetitions and exposures) and the “automatic” way it functions when implicated in action (Brownstein 2018, chap 3). In addition, transforming the property from implicit to explicit changes the nature of the entity; in fact, this “explicitation work” is part of the way in which people can begin to remove (unwanted) implicit attitudes, implying that this property is essential to the existence of the entity as such (Bownstein 2018, chap 7).
This differs from implicitness when it comes to public cultural kinds. Complex classification systems (on paper) have both implicit and explicit aspects, but it is an ontological mistake to speak of implicitness as a holistic (definitional) property of the system. This leads inexorably to “mentalistic” (and thus ontologically problematic) conceptualizations of what should be non-mental cultural kinds. As we have seen, mentalistic conceptualizations of the implicit aspects of public cultural kinds (e.g., “tacit collective” presuppositions, worldviews, assumptions, etc.) brings with it all the problems of the “collective mind” tradition that began with Wundt and went through Durkheim and onwards to contemporary heirs (Turner 1994, 2014). This is something to be avoided if we can.
Implicitness in public cultural kinds should instead be seen as an incidental, observer-relative, property because it can be changed without thereby altering the nature of the entity (and, as we have seen, depends on the observer, a point that will have to be developed in a future post). For instance, if we take a text corpus and make a lot of its implicit features explicit (via hand-coding, computational text analysis, and other methods), it still remains the public cultural kind that it is (a collection of texts). Taking tacit personal knowledge and making it explicit personal knowledge is, by way of contrast, if not wholly impossible, only feasible by radically transforming the original cultural-cognitive kind into something else, as Polanyi noted long ago. This last implies an ontological shift, the first case does not.
Implicit personal and implicit public are thus two distinct, but conceptually related properties of different types of cultural kinds.
Arseniev-Koehler, A., & Foster, J. G. (2020). Machine learning as a model for cultural learning: Teaching an algorithm what it means to be fat. https://doi.org/10.31235/osf.io/c9yj3
Bidney, D. (1968). Theoretical Anthropology. Transaction Publishers.
Biernacki, R. (2000). Language and the Shift from Signs to Practices in Cultural Inquiry. History and Theory, 39(3), 289–310.
Brownstein, M. (2018). The Implicit Mind: Cognitive Architecture, the Self, and Ethics. Oxford University Press.
D’Andrade, R. (1991). The identification of schemas in naturalistic data. In M. J. Horowitz (Ed.), Person schemas and maladaptive interpersonal patterns (pp. 279–301). University of Chicago Press.
Karmiloff-Smith, A. (1995). Beyond Modularity: A Developmental Perspective on Cognitive Science. MIT Press.
Lizardo, O. (2017). Improving Cultural Analysis: Considering Personal Culture in its Declarative and Nondeclarative Modes. American Sociological Review, 82(1), 88–115.
Lizardo, O., & Strand, M. (2010). Skills, toolkits, contexts and institutions: Clarifying the relationship between different approaches to cognition in cultural sociology. Poetics , 38(2), 205–228.
Rinaldo, R., & Guhin, J. (2019). How and Why Interviews Work: Ethnographic Interviews and Meso-level Public Culture. https://doi.org/10.31235/osf.io/87n34
Rotolo, M. (2020). Culture Beneath Discourse: An Ontology of Cognitive Cultural Entities. https://osf.io/preprints/socarxiv/v39te/
Sewell, W. H., Jr. (2005). The concept (s) of culture. Practicing History: New Directions in Historical Writing after the Linguistic Turn, 76–95.
Sperber, D. (1996). Explaining culture: A naturalistic approach. Blackwell Publishers.
Swidler, A. (2001). What anchors cultural practices. In K. K. Cetina, T. R. Schatzki, & E. von Savigny (Eds.), The Practice Turn in Contemporary Theory (pp. 74–92). Routledge.
Turner, S. P. (1994). The Social Theory of Practices: Tradition, Tacit Knowledge, and Presuppositions. University of Chicago Press.
Turner, S. P. (2007). Practice Then and Now. Human Affairs , 17(2), 375.
Wuthnow, R., & Witten, M. (1988). New directions in the study of culture. Annual Review of Sociology, 14(1), 49–67.