The impossible mind of sociology

The mind has not been a central concept in sociology. According to the traditional view, the mind is located in the brain, and is thus bereft of observable social facts for sociological studies. At most, it is a concept of psychology or philosophy. This article argues that the history of the modern novel provides large amounts of data about minds and consciousness. Even though individual novels are fictional and invented, the continual reception of these fictional presentations verifies their social relevance. The article argues that fiction establishes the main social discourse on possible private thoughts, thus having a great impact on how we understand and speak about minds and human interiority. The argument is advanced by selectively reading a long-standing narratological debate on literary minds and their exceptionality. The article renounces the cognitive theories of ‘mind-reading’ as overly optimistic and metaphorically misleading, resorting instead to the phenomenological theories of ‘primary intersubjectivity’, which help in understanding how novelists are able to invent credible minds in the first place.

vados posibles, por lo cual tiene una gran repercusión en cómo entendemos y hablamos de las mentes y de la interioridad humana. El argumento va más allá de la lectura selectiva de un debate narratológico antiguo sobre las mentes literarias y su excepcionalidad. El artículo renuncia a las teorías cognitivas de "lectura de la mente" por ser demasiado optimistas y por inducir metafóricamente al error. En cambio, recurre a las teorías fenomenológicas de "intersubjectividad primaria", que, sobre todo, ayudan a entender cómo los novelistas son capaces de inventar mentes creíbles. mente, novela, excepcionalidad, narratología, lectura de la mente, intersubjectividad primaria For many reasons, the mind has seldom been profiled as a sociologically significant concept. The mind appears to belong to psychology or philosophy, and it is thus bereft of socially observable facts for sociologists to study. Most problematically, the mind tends to be located within the skull -i.e., in the brain -and thus beyond the reach of the sociological gaze. I hold a more expansive conception of the mind and argue that literary fiction -the novel and short story in particular -has provided a socially observable and historically attested discourse of the mind. The long history of reading fiction proves the validity of these fictional minds and transforms the general understanding of the potentialities of other minds among readers. My method in this article is to selectively re-read a recent debate on fictional minds and their exceptionality in literary theory by focusing on the historical conditions for the possibility of the emergence of exceptional and textually available third-person minds. I intend to tease out some relevant particularities of the fictional discourse on minds with the help of Ian McEwan's novel The Children Act (2014).

Palabras clave
At first sight, it may seem odd to turn to literary fiction and literary theorists when reflecting on issues concerning minds and consciousness. I want to emphasise that there is no need to assume that novelists in general are especially profound theorists of the mind or consciousness. What is obvious, however, and what I elaborate further below, is that many novelists have indeed engaged in nuanced games with various minds, often with minds in the business of trying to understand and communicate with other minds. To put it bluntly, I am less interested in the thoughts and ideas novelists display as such than I am in the special practices and operations that fiction enables, especially novels as discursive sites for playing with minds. From this perspective, fiction can be understood as a historically evolving practice of language use with particular linguistic and rhetorical affordances in portraying, elaborating, and analysing minds. As I will argue in this article, fiction is an exceptional site for experimenting with minds and mind attributions. I propose the use of these resources within fiction generates sociologically relevant facts, namely discursive presentations of other minds and their working in context.

The distinctiveness of fiction
At least since the seminal work of Dorrit Cohn (1978;1999), the issue of (re)presenting minds has been profiled as one of the most crucial distinctions between literary fiction and non-fiction. Cohn was an ambitious writer. The titles of her books Transparent minds: Narrative Modes for Presenting Consciousness in Fiction (1978) and The Distinction of Fiction (1999) finely portray the particular relevance of fictional minds for the distinctiveness of the whole genre of narrative fiction. In brief, Cohn's argument is that we are able to know the most intimate thoughts of fictional minds in ways that remain impossible with real-life minds.
How can we justify such an exceptionality? According to Cohn and the narratological convention, fiction writers do not represent other minds; they present them: "In depicting the inner life, the novelist is truly a fabricator" (Cohn 1999, p. 6). The novelist is the creator of his or her characters, and because of this creation, the novelist is always able to describe even the deepest thought. "[A] character […] is known to his narrator in a manner no real person can be known to a real speaker" (Cohn 1999, p. 117).
When Cohn (1999) discusses the 'signposts' of fictionality and the distinctiveness of fiction, it is essential to recognise the frame and direction of her polemic. Once and again, she accentuates the clear and pertinent distinction between fiction and historiography. This is also the case when Cohn (1999, p. 117) begins with a quote from the philosopher John Searle who suggests that "There is no textual property, syntactic or semantic, that will identify a text as a work of fiction". Searle defended his bold claim with a reference to the beginning of Iris Murdock's novel The Red and the Green. Cohn points out that this example effectively testifies against Searle's own case, because "what 'serious' discourse ever quoted the thoughts of a person other than the speaker's own?" As for history, Cohn's claim is apposite and highly relevant, for example, in discriminating against the pan-fictional theories of Hayden White (1978;1987). Following his formalist ideas, White famously argued that because the novel and historiography both rely on narrative form, they are, for this reason, intrinsically and equally fictional. Cohn reminds her readers of several other authors (e.g. E. L. Doctorow and Arnold Toynbee) who also reason that the category of 'narrative' as such eliminates the difference between fiction and non-fiction. Within this debate, it is easy to affirm Cohn's position, as she insists on the essential formal differences between different narrative genres.
However, Cohn's term 'serious discourse' engenders some further ambiguity. The expression is vague, rendering the argument seriously open and simultaneously highly normative, downplaying the relevance of all possible deviations. To contextualise the comment, it is useful to remember that prominent narrative theorists of the time such as Paul Ricoeur (1984;, White, and Cohn all share one important and problematic point of departure in theorising narrative. They all focus on comparing history and literature as the dominant genres of narrative, largely disregarding the whole existence of everyday oral narration, as studied and theorised within the thriving sociolinguistic tradition of William Labov and Joshua Waletzky (1997). 1 Does the suggested idea of non-serious narrative discourse refer to these oral versions? I return to the consequences of this problem later, but I will remark here that Mari Hatavara and Jarmila Mildorf (2017) have recently documented how narrators in vicarious storytelling indeed keep accounting for other people's thoughts and mind contents (see also Browse and Hatavara, forthcoming).
Cohn's argument for the particularity of fiction in presenting third-person minds is powerful, yet possibly too categorical. If, as Käte Hamburger (1993, p. 83) points out and Cohn affirmatively states, "Epic fiction is the sole epistemological instance where the I-originarity (or subjectivity) of a third person qua third person can be portrayed", and if the novelists 'know' the characters' minds only because they have themselves created them, at least one confusing problem still remains. This argument does not include any elements that could help us to explain the birth and evolution of this aptitude to present elaborate and credible fictional minds. How, indeed, has this exceptional capacity emerged exclusively within fiction?
This problem is in parallel with another issue often used in arguing for the exceptionality of fiction, namely the question of referentiality. If my argument is correct, in the following passage Cohn is still drawing a distinction between historiography and fiction, as she maintains, "The adjective nonreferential in the definitional phrase 'nonreferential narrative' needs to be qualified at somewhat greater length. First and foremost, it signifies "that a work of fiction itself creates the world to which it refers by referring to it" (Cohn 1999, p. 13, italics added). Historians are not allowed to create their own worlds without a point-bypoint reference to archival documents. This is not to say that everything in historiography is factual or that historians do not try to pursue their own agendas and tell their own versions of the past. Nevertheless, historians cannot elaborate the contents of historical minds without clear, textual sources. Historians are indeed liable to explain that the past world they describe is not of their own making.
Of course, the problem of referentiality is not quite so straightforward, as Cohn herself is the first to admit with an example from Flaubert's Sentimental Education. Novels are able to create double frames, Cohn argues, setting first a historical, factual, and referential frame, and then a fictional frame with fictional characters within it. After all, we can accept that Don DeLillo's (2007) Falling Man refers to the real Manhattan and the events of 9/11. A similar problem arises with geography. For example, in McEwan's (2014) The Children Act, the actual geography of London, together with the invented places, does not merely constitute a passive context for the events to happen; it plays an active role inside the fictional narration (Hyvärinen, 2018). Gradually, the distinction between referential and nonreferential genres rather begins to resemble the difference between Weberian ideal types rather than an empirically valid distinction. The beginning of Hilary Mantel's author note in A Place of Greater Safety provides a good example: This is a novel about the French Revolution. Almost all the characters in it are real people and it is closely tied to historical facts -as far as those facts are agreed, which isn't really very far. It is not an overview or a complete account of the Revolution. (Mantel 2010, p. ix) In short, the difference between 'referential' and 'non-referential' seems to be a difference of degree, not of kind, at least for the novelists themselves. With minds, the problem of 'the nonreferential' cuts even deeper. Let me repeat Cohn's (1999, p. 13) prime example of non-referential fiction, taken from Franz Kafka's The Castle: It was late in the evening when K. arrived. The village was deep in snow. The Castle hill was hidden, veiled in mist and darkness, nor was there a glimmer of light to show that a castle was there. (Kafka, 1969, p. 3) It is not difficult to observe that this opening passage and the whole novel is much more deeply non-referential than Hilary Mantel's work. We may want to afford a particular aesthetic value to Kafka's work just because of this profound non-referentiality -or perhaps not -but aesthetic evaluation is not my primary concern here. For a European reader, at least, such words as 'village', 'castle', 'mist', and 'darkness' already have a distinct referential value on a more generic, cultural level. We most likely do not understand castles as children's playgrounds, but attach to them connotations of history, authority, privilege, and possibly detention. As much as the characters, events, and thoughts are invented and onerically non-referential, the novel powerfully builds on what can be called generic cultural referentiality. The novel is not 'about' a particular event, like Mantel's novel; nevertheless, it is 'about' something culturally recognisable. As much as the level of events and characters is non-referential and invented, the cultural level of darkness is referred to and vividly evoked. David Herman (2002, p. 101) experiments with the absolute limits of narrativity, giving at the same time good examples of texts utterly devoid of referentiality, such as "A sluppa fiblo. Sim a gingy beeble the yuck I the splubba orbia". If everything in fiction were invented and non-referential, most of us would not be keen on reading novels, since they would not speak or matter to us in any significant way.
In a recent, much debated contribution to the problem of fictionality, Henrik Skov Nielsen, James Phelan, and Richard Walsh (2015, p. 62) similarly emphasise that "the use of fictionality is not a turning away from the actual world but a specific communicative strategy within some context in that world". However, soon afterwards they instead maintain that the "ability to invent, imagine, and communicate without claiming to refer to the actual is a fundamental cognitive skill, one crucial to humans' interactions with their world and their fellow beings in that world" (Skov Nielsen, Phelan and Walsh 2015, p. 63, italics added). According to my argument, this categorical, unequivocal rejection of referentiality is a poor and ambiguous criterion for fictionality. To escape this confusing wavering -to turn or not to turn away from the actual world -I suggest a new distinction between event-historical referentiality and broader cultural referentiality. Fiction too refers to the actual world, but its methods of reference are different, more complex, and indirect compared to those of non-fiction.
The idea of there being two different kinds of referentiality becomes more salient when we return specifically to minds. The standard, categorical claim that the novelists do not 'represent' the characters' minds because they fictionally 'create' them becomes somewhat problematic when we introduce this notion. Readers already know -at least distantly -castles, villages, and darkness, but according to Cohn's own theory, no such knowledge of other minds exists outside fiction. If other people really are such black boxes whose minds are utterly inaccessible to us, we run into two problems. Firstly, from where do the novelists draw their 2. Cognitive narratology is a much wider and richer project than is possible to cover here. Monika Fludernik (1996), David Herman (2002;), Manfred Jahn (2005, and Bernaerts et al. (2013) are just a few influential texts within this approach. Patrick Colm Hogan (2003, p. 240), for example, does not share the ideas of mind-reading and emphasises instead "We do not have access to anyone's experiential subjectivity. We cannot feel anyone else's emotions. Thus, to identify their emotions, we must rely on the clues from what can experience. Prototypes are extremely helpful in this regard". credible insights about minds? Secondly, if the imagined fictional minds are entirely non-referential, residing outside the cultural layer of referentiality, how can we explain the readers' interest in encountering these minds and finally even understanding these entirely invented entities?

Mind-reading as a solution?
Around the turn of the millennium, literary narratology experienced a significant theoretical rejuvenation and boom in new theories (Fludernik, 2005). This major change of agendas, methods, and theories was quickly characterised as a move from classical (structuralist, formalist) narratology to postclassical (cognitive, rhetorical, unnatural) narratology (Herman 1999a;Alber and Fludernik 2010). "Cognitive narratology can be defined as the study of mind-relevant aspects of storytelling practices" and as "a subdomain within 'postclassical' narratology" (Herman, 2009, p. 3). The approach draws heavily on cognitive science, cognitive psychology, and the neurosciences; it applies their theories and concepts in the analysis of fiction, often with the mission of making literary scholarship more scientific. Some representatives of cognitive narratology 2 -most prominently Lisa Zunshine (2006) and Alan Palmer (2004; -radically contest Cohn's basic assumptions about the accessibility of everyday minds. At the same time, the whole frame of discussion in narratology changes, since cognitive narratologists are not primarily interested in drawing a line between history and fiction. Rather, they focus on elaborating the continuities between fiction and everyday experience. The title of Lisa Zunshine's (2006) influential book, Why We Read Fiction: Theory of Mind and the Novel, summarises the crucial argument. Theory of Mind (ToM), which was developed within cognitive science, is a central element both in Zunshine's and Palmer's theories of fiction. While Cohn argued for the distinctiveness of fiction and narratology as a distinct theory about fiction, cognitive narratologists often celebrate the achievements of cognitive science and cognitive psychology as general theories for building narratology and fiction studies. David Herman (2001, p. 2) outlines this new hierarchy of disciplines by suggesting "both narrative theory and linguistics should […] be construed as resources for cognitive science". In his later work, Herman (2011a) rejects this hierarchy, but Alan Palmer (2011, p. 200), for example, endorses the position with a far more radical formulation (see below).
Zunshine uses the terms 'mind-reading' and 'Theory of Mind' almost synonymously. She employs the latter concept "to describe our ability to explain people's behaviour in terms of their thoughts, feelings, beliefs, and desires" (Zunshine 2006, p. 6). This is indeed vital mental vocabulary in accounting for human action, but her description of ToM nevertheless contains at least two pertinent shortcomings. Zunshine does not properly define or explicate the methods and resources for obtaining adequate input information about other minds, which, problematically enough, she seems to explain in terms of the parallel concept, 'mind-reading'. However, the command of mental vocabulary does not say much about the problematic capability of detecting thoughts. In fiction, the behaviour of the characters can of course be explained in terms of "their thoughts, feelings, beliefs, and desires", because the author 'knows' the characters in a way we seldom know people in real life. Furthermore, we can presume that Zunshine merely uses the term 'mind-reading' as a broad metaphor, yet this metaphor promises unspecified access to other minds and effectively effaces the crucial difference between reading fiction and observing reallife people from afar. Do we indeed understand others in real life by being familiar with the intimate contents of their thought, as the metaphor suggests?
As the philosopher Daniel D. Hutto highlights, the word 'theory' in ToM promises too much and gravely misrepresents the ways we understand others. Understanding and competently using the vocabulary of intentions is far from using a theory, at least a theory providing us with cogent information about other peoples' mind contents. Knowing the ambiguity and volatility of feelings, desires, and beliefs, the computational machine needed for making sensible conclusions should be enormous and amazingly quick at the same time (on this criticism of ToM, see Hutto, 2004;. In describing the acquisition of ToM by young children, Zunshine (2006, p. 8) recounts the often reported 'false belief test': In one of the more widespread versions of the test, children see that 'Sally' puts a marble in one place and then exits the room. In her absence, 'Anne' comes in, puts the marble in a different place, and leaves. Children are then asked, "Where will Sally look for her marble when she returns?" Small children think that Sally will look for the marble in the place Anne has put it, not understanding that Sally does not know that Anne has changed the hiding place. The test obviously documents one important stage in human cognitive development. Yet it says preciously little about having a command of any finer set of mental concepts, or about how children understand complex reasons, desires, and beliefs. Understanding that people sometimes have false beliefs is some way away from knowing that in this particular situation and issue, this particular person may have an entirely false belief. A vast storage of cultural knowledge, rather than a mere psychological device, is needed to understand the complex networks of false beliefs.
The fundamental problem with the test is that it frames the folk psychological understanding of minds in terms of a 'spectator sport', as Hutto (2004) puts it. In real-life situations, we do not sit behind a glass window and try to 'read' other unknown minds from afar. Infamously, paranoia relies on a tendency to "read" and misunderstand minds from afar even without interaction (Hyvärinen, 2015). If the understanding of other minds was based on such an abstract and general cognitive device that gave us broad access to 'read' other minds, there would not be much left of the particularity of fiction that Cohn theorised.
Neither Zunshine nor any other cognitivist theorist claims that ToM provides limitless or unerring access to other minds. On the contrary, Zunshine (2004, p. 13) openly admits that "the process of attributing thoughts, beliefs, and desires to other people may lead to misinterpreting those thoughts, beliefs and desires". However, cognitive narratologists have remarkably little to say about why ToM and mind-reading fail or succeed in particular situations, which renders the idea of mind-reading rather pointless in empirical analysis. This instability of interpretation, Zunshine argues, may, however, be one of the reasons why we read fiction: It is possible, then, that certain cultural artefacts, such as novels, test the functioning of our cognitive adaptations for mind-reading while keeping us pleasantly aware that the 'test' is proceeding quite smoothly. (Zunshine 2004, 18, p. italics added) Playing with different minds and tracking down the thinking and feeling of diverse minds in novels may indeed be one of the literary games that fascinate readers, yet Zunshine's formulation threatens to reduce the entire complex literary experience to an instrumental exercise in the service of the evolutionary skills of mind-reading. Here, Zunshine rejects the exceptionality of fiction proposed by Cohn and postulates it as a mere auxiliary of real-life mind-reading. Nevertheless, Zunshine succeeds perfectly in documenting and elaborating the readers' fascination with the most varied fictional minds. Perhaps we could now re-formulate her title to ask why we read fictional minds. The answer, I suggest, should include both everyday attempts at understanding other people and Cohn's claim about the particularity of fictional minds.

The critics of mind-reading
Cognitive narratology was designed to render literary studies more rigorous and scientific. For a sociologist like myself, a striking feature of its notions of mind and mind-reading is the complete absence of social and interactional aspects. It assumes that people have simply become capable of reading other minds thanks to evolution and cognitive development. Dan Hutto and Shaun Gallagher (2007; in particular discredit this model. Gallagher points out that even small children are able to follow and understand the intentional, emotional states of adults. This understanding is embodied and immediate, and it is surely not mediated by any theory apparatus running in the child's mind. Gallagher (2009, p. 293) also notes that infants 'are not taking an observational stance; they are interacting with others. For example, infants vocalise and gesture in ways that are affectively and temporarily 'tuned' to the vocalisations and gestures of the other person'. Gallagher calls this phenomenon "primary intersubjectivity" and reminds us that it "is not something that we leave behind as we mature. We continue to rely on our perceptual access to the other's affective expressions, the intonation of her voice, the posture and style of movement involved in her action". This understanding is not based on abstract theory; it can accumulate with the help of intersubjective experience, and it has nothing to do with 'reading' other people's thought contents.
Based on such observations, Gallagher rejects what he calls "the Cartesian idea that other minds are hidden away and inaccessible", and suggests instead that "we directly perceive the other person's intentions, emotions, and dispositions in their embodied behaviour" (2009, p. 292). Gallagher steps determinedly out of the conceptual scheme of mind-reading by accentuating that our "normal, everyday stance toward the other person is not third-person, detached observation; it is secondperson interaction". Gallagher's 'primary intersubjectivity' thus transfers the issue of the mind to an intersubjective level. As regards literary minds, it helps to understand how novelists are able to invent credible minds without claiming that we could read some textual mind contents in third-person relationships in the way we read them in novels. Gallagher's primary intersubjectivity is a strictly defined, limited phenomenon without a hint of accessing other people's conceptual thought contents, as mindreading keeps doing.
Cognitivists typically call the command of ToM "folk psychology". In such use, the term is rather confusing because there is normally not much 'folk' in cognitivist theories, since they build on intrapsychic capacities. Hutto instead draws on Jerome Bruner's entirely different idea of folk psychology. Bruner, a disenchanted former cognitive psychologist himself and a pioneer of Vygotskyan cultural psychology, locates the issues of folk psychology and understanding others at the social level (Bruner, 1990;Amsterdam and Bruner, 2000). "Folk psychology […] is a culture's account of what makes human beings tick", he argues (Bruner 1990, p. 13). Bruner's folk psychology is based on the cultural knowledge of canonicity, on cultural and normative expectations about how things should be and come about within a particular culture (1990, p. 39-40).
For Bruner, folk psychology is far from a deep psychological gaze into other minds and thoughts. Instead, it denotes our local cultural knowledge resources. Most of the time, we understand what other people do and can do because we understand the cultural canonicities, frames, and scripts. Of course, the exceptional moments and deviations from expectations are critical moments in understanding the minds of others. ToM theorists suggest that in cases of such deviations, people start their ToM devices in order to make sense of the thinking of others. However, Bruner also frames this situation in social terms. He maintains that "it is only when constituent beliefs in a folk psychology are violated that narratives are constructed" (Bruner 1990, p. 39). Where ToM theorists see atomistic individuals running a theory, Bruner looks at the social situation and sees the participants' need to give and order narrative accounts of deviant behaviour. Hutto (2008, p. 46) argues that asking "why did you do it this way?" or any other version of a "why" question in interaction is a much more powerful and reliable method than immersing oneself in speculative mind-reading.
Gallagher and Hutto do not entirely refute the existence of the processes described in terms of ToM, but they argue that those processes are rather marginal and sporadic. On the one hand, understanding others is fairly immediate, interactional, and embodied rather than being theory-driven. On the other hand, it is based on a command of cultural knowledge and an achieved familiarity with the people to be understood. When I have a light conversation on domestic issues with my wife, I often know her responses in advance, even down to the words she will choose. Following Gallagher, the insight is immediate and based on a history of an interactionally shared field of humour. However, without shared prompts or ongoing conversation, I cannot claim any capacity to know the contents of her thought. Gallagher and Hutto therefore help us to understand the successes and failures in understanding other minds, because they outline the crucial resources for understanding others more concretely when compared to the ToM theorists. As they argue, the growth of understanding is built on interaction, both on telling folk psychological explanatory narratives and on asking the decisive 'why' questions when expectations have been breached. A lack of familiarity, inadequate cultural resources, and missing interactions tend to generate misunderstanding and speculative mind-reading.

Alan Palmer and the extended mind
The literary theorist Alan Palmer (2004Palmer ( , 2010 defends a markedly externalist view on fictional and real-life minds. 3 He does so by often taking a somewhat radical cognitivist position. When Cohn (1978) wrote of fiction's 'transparent minds', another literary theorist, Ann Banfield (1982, 211), aptly criticised the choice of terms by saying that represented consciousness "does not create 'transparent minds'. The mind is never transparent, not even to 'omniscient narrators'". In Fictional Minds, Palmer (2004, p. 132) initially takes an entirely different position: I wish to argue here that not only can fictional minds be transparent to readers, there is a strong sense in which real minds can be transparent to other people. Daniel Dennett makes the point that your 'body can vigorously betray the secrets you are desperately trying to keep… 'Transparent' is arguably not the most apposite term to characterise minds. People would not make the amount of wrong choices they do if they had a transparent vision of their own and other people's minds. The problem with perceiving minds is partly due to their volatile and inexhaustible nature. Quite correctly, Palmer notes that in fiction other people often better recognise the thoughts of a character than the character does him-or herself. Such situations would of course be impossible in real life if we truly had a transparent view of our own minds. Language can never exhaustively communicate real-life experiences (e.g. Brockmeier, 2008). In the case of fiction, there is of course only the words on the page, but rather than providing clear-cut and transparently dull minds, artful fiction alludes to the same inexhaustible nature of minds with all its layers, niches, and allusions. Six years later, in Social Minds in the Novel, Palmer's (2010, p. 44) terminology and position has changed accordingly: An emphasis on social minds will inevitably question these twin assumptions: first, that the workings of our own minds are never accessible to others; and, second, that the workings of our own minds are always and unproblematically accessible to ourselves.
While transparency is primarily a categorical yes-or-no quality, accessibility affords more flexibility and space for negotiation. The contents of a real-life mind are to some extent accessible, but 3. Palmer introduces a remarkable amount of new discussions outside literary theory; a contribution I cannot cover fully in this article. In discussing the extended mind, Palmer (2010, p. 51) also introduces such authors as Andy Clark, who "are also interested in another aspect of the mind that is called physically distributed cognition: 'our habit of off-loading as much as possible of our cognitive tasks into the environment itself'" (Dennett 1996). Andy Clark calls this process "Supersizing the Mind" (2009). 4. Scholars may appear to write their academic work in solitude, but they are constantly in discussion with other scholars through their contributions. this is not the same thing as claiming that somebody has actually accessed them, or, for that matter, that minds are accessible in every last detail. Palmer correctly notes that even in fiction, there are minds that resist the reader's attempts to access them.
Palmer emphatically promotes the concept of the social mind to depict "those aspects of the whole mind that are revealed through the externalist perspectives" (Palmer, 2010, p. 39). Next, he proceeds to outline his most original contribution: "An important part of the social mind is our capacity for intermental thought. Such thinking is joint, group, shared, or collective, as opposed to intramental, or individual or private thought" (Palmer, 2010, p. 41, italics in the original). According to Palmer, literary scholars have so far focused too one-sidedly on characters' thinking in solitude and have passed over the workings of the mind during action and interaction. Who is indeed living his or her life by listening to inner monologues? Palmer (2010, p. 42) asks "Why assume that the self can only be found (or easily found) in solitude?" 4 The mind beyond the skin should replace the old, limited ToM within the skull. Minds should not be automatically located 'between the ears', as it is frequently put. Therefore, minds should not escape the sociological gaze either.
To refine his theory, Palmer (2010, p. 46-48) proposes a typology of social minds in novels, thereby introducing a whole set of sociological perspectives to fiction. He suggests the categories of intermental encounters: small intermental units (e.g. families), medium-sized intermental units (e.g. colleagues, neighbours), large intermental units (towns and villages), and intermental minds (tightly consolidated units of any size). For a sociologist, this categorisation offers mind-related phenomena at several levels of social organisation, even though it appears in current form somewhat ad hoc and devoid of a systematic theoretical grounding. Palmer (2010, p. 46) explains "intermental encounters" by saying that it "is not possible to have a coherent dialogue without at least some intermental communication". The psycholinguist Stephen C. Levinson (2006, p. 45) similarly argues that "the interpretation of others" behaviour is a precondition for interaction', and suggests thus that humans have an "interaction engine" before actual interaction. The metaphor is typically cognitive, and refers to phenomena that Gallagher discussed in terms of "primary intersubjectivity" and adaptive intentionality.
By "small intermental units", Palmer refers to "small groups of various sorts such as marriages, close friendships, and nuclear families" (2010, p. 47). Here, the participants typically have lots of experiential knowledge of how the others behave, think, and Digithum, No. 24 (July 2019) | ISSN 1575-2275 A scientific e-journal coedited by UOC and UdeA 43 Matti Hyvärinen, 2019 FUOC, 2019 express themselves. Without entering into any sort of mindreading, couples, for example, may also have a huge store of local canonicity, an understanding of how things normally are. In McEwan's The Children Act, the protagonist -Fiona Mayephones Jack, her husband, while on a business trip. She shows how well she is able to read the signs because of the familiarity of the setting: When she heard his tentative hello, the acoustic told her that he was in the kitchen. The radio was playing, Poulenc perhaps. On Saturday mornings they always had, always used to have, a lazy but early breakfast, a spread of papers, muted Radio Three, coffee, warmed pain aux raisins from Lamb's Conduit Street. He would be in his paisley silk dressing gown. Unshaven, hair uncombed. (McEwan, 2014, p. 181) Before Jack says anything more than hello, Fiona has located him within the kitchen and their Saturday morning routine. She has done all this without any indication of reading Jack's mind. It is also noteworthy that various material elements (e.g. the acoustics, the radio, and warmed pain aux raisins) participate in this emerging thinking beyond the skin. The correction of the expression ("they always had, always used to have" -the couple are experiencing marital difficulties) indicates that Fiona is indeed doing the thinking here.
By "medium-sized intermental units", Palmer refers to units "such as work colleagues, networks of friendships, and neighbourhoods". What is different from the second category, Palmer explains, is that now 'the emphasis is less on individuals knowing what another person is thinking and more on people thinking the same way' (Palmer, 2010, p. 48). More precisely, we are talking about shared cultural knowledge the actors are able to mobilise in different situations. In order to elaborate what this could mean, I return to The Children Act, just before the passage quoted above. By mistake, possibly because of her confused motivations, Fiona -a judge -ends up kissing an 18-year-old boy, the subject of a law case over which she presided. The next day, while sitting in a car with Paulig, her assistant, Fiona ponders: She was not prone to wild impulses and she didn't understand her own behaviour. She realised there was much more to confront in her confused mix of feelings, but for now it was the horror of what might have come about, the ludicrous and shameful transgression of professional ethics, that occupied her. (McEwan, 2014, p. 180) This short passage aptly summarises several themes discussed previously. We see, for instance, that Fiona's mind was anything but transparent to herself or to the reader. Sitting in the same car, Paulig is possibly able to notice Fiona's anxiety, but there are no available ways he could 'read' Fiona's thoughts since the kiss never became public. Without experiential knowledge and interaction, the idea of mind-reading falls short in terms of understanding the contents of thought. The abstract theory of mind-reading can only notice that "Oh, it seems that Paulig fails in his mindreading". In precisely the way Palmer suggests, Fiona makes her profession and its collective thinking relevant in her inner mind by thinking about her "ludicrous and shameful transgression of professional ethics", thus inviting the disapproving tone of the assumed collective response. In a manner of speaking, she is not thinking alone, "within her skull"; she has activated the voice of the judgmental, professional chorus as part of her worrying. A few lines below, Fiona's mental torture continues: Even now, miles behind her in London, the case was being discussed. That one day soon she'd hear on her phone the hesitant embarrassed voice of a senior colleague. Ah, Fiona, look, awfully sorry but I'm afraid I should warn you, uh, something's come up. Then, waiting for her back at Gray's Inn, a formal letter from the Judicial Complaints investigation officer. (McEwan, 2014, p. 181) Here again, only the readers are able to read these thoughts. Fiona's shame is expressed in the typical form of hypothetical narration; she pre-figures her colleagues' reprehension in her mind (Karttunen, 2015). Because of her experience and familiarity with the professional ethics and manners of her colleagues, Fiona is able to envision in a detailed, fictionalised way the prospective embarrassment and words of her colleagues. She knows the rules, the ethics, the manners, and the canonical course of events. In the novel, the judges recurrently disagree on the decisions made by other judges, yet this collectively shared and institutionalised layer of professional culture actively participates in Fiona's thought. In her imagination, Fiona is able to read the public, canonical mind of the profession precisely because it consists of public and shared knowledge. However, there is no big juridical mind doing this thinking, since Fiona herself thinks and evokes the canonical expectations of her profession. Seen from another angle, Fiona is not only thinking individually; she also animates the shared thoughts of her profession. When Fiona arrives at the idea of the official letter waiting for her at Gray's Inn, she finishes her dark thoughts, takes her phone and calls Jack. Together with the imagined official document and her phone, Fiona crosses the boundaries. Passing from the cultural and the canonical level, she returns to her home and the sphere of intensive interaction (however difficult due to her crises). These passages challenge the ideas of the transparency of minds and the unqualified capacity of mind-reading. They also register the wider social resources of understanding minds.
In his discussion, Palmer succeeds in convincing how minds and mind processes are not confined within the skull or inside individual brains. His linguistic strategy of choice, of naming 'minds' of different levels of social organisation nevertheless evokes questions. The idea of various minds as nouns runs the risk of encouraging ideas about such minds being self-enclosed, stable, and semi-organic entities. The issue of the mind's sociality might equally be addressed in terms of networks of meaning making, thus focusing on the social and spatial distribution of thinking, remembering, and feeling, without establishing new layers of collective organs.
Palmer's cognitive reductionism may further establish problems for his attempt to theorise the mind's social aspects. Not only does he draw on resources and results from cognitive science; he makes them the essential basis of his theory: "To talk of a cognitive approach to literature can be rather misleading if it gives the impression that it is simply one alternative among a range of others… the cognitive approach is the basis of all the others" (Palmer 2011, p. 200). His view on fiction similarly threatens to remain narrowly reductionist, as he claims, "Fictional narrative is, in essence, the presentation of mental functioning" (2011, p. 202, italics added). These formulations, alas, render the bulk of the resources of social sciences and literary theory marginal. Palmer's attempt to understand and thematise the mind beyond the skin is refreshing and welcome, yet the goal is hard to realise without renouncing the cognitive reductionism that eventually keeps privileging the processes within the skull. As one of the leading cognitivist theorists puts it, "Despite Palmer's assertion, cognitive science offers no help here. If we follow the standard neuro-cognitive view that the mind is a function of the brain, then there has to be a brain for there to be a thought" (Colm Hogan, 2011, p. 244).

The social relevance of fictional minds
The core sociological relevance of fictional minds does not primarily ensue from the representational truthfulness of these minds. As the literary theorist Maria Mäkelä (forthcoming) states, 'writing generates interiority, and not vice versa'. Arguably, there is traffic in both directions, but my interest in this paper is in the way literary genres generate nuanced discursive tools for discussing the working of minds. There is no other institutional form of discourse that could present and study human emotions, thoughts, bodily sensations, talk, and behaviour as a holistic process. Patrick Colm Hogan (2003, p. 1) aptly remarks that when "empirical researchers in the social sciences consider the nature of emotions and emotion concepts […] with only a few exceptions, they almost entirely ignore a vast body of existing data that bears directly on feelings and ideas of feelings -literature, especially literary narrative". The existence of this established institution and its experiments indicates that the understanding of what is "empirical" badly needs rethinking in sociology.
By rephrasing the words of Marina Grishakova (2014, p. 7), "fictional narratives work as experiential labs where various hypotheses and inferences about the functioning of the [mind] are imaginatively tested". The history of the novel is accordingly partly about teasing out and communicating aspects of the human mind. "However, fictional narratives reach consequences not through the austere logic of argumentation, but by throwing in and elaborating on supervening details, displaying new circumstances and alternative paths" (Grishakova 2013, p. 7). In experimenting with connections between the body, emotion, thought, talk, and behaviour, fiction is significantly a theoretical project that should not be bypassed by considering it merely from the perspective of lacking referentiality. This is all sociologically relevant, since there is no alternative discursive field in which to experiment so systematically with possible private thoughts and processes. First the epistolary novel and then the modern novel not only expressed the feelings and thoughts of their protagonists; they effectively generated the language, metaphors, and discourses to express and refine human interiority. Due to the continuing reception of these fictional works and their role in education and even modern nation-building (Moretti, 1998), we have reasons to re-evaluate their sociological relevance.
Different accounts about the history of the novel and the mind seem to verify one general observation. There is no linear history of minds in novels, and there is a lot of contingency in the ways of attributing the contents and processes of the mind to characters (see the contributions in Herman, 2011a). Such particularities include the relevance of the landscape, weather, and physical environment in the Romantic era (Vallins, 2011), as well as the mid-nineteenth-century exploration of human physiology as an indicator of the mind (Dames, 2011). Novelists do not simply approach people, 'read their minds', and write about their thoughts; they have flourishing literary genres with related affordances and all the non-literary, popular, and scholarly debates to draw on. For example, Anna Burns' Milkman -a recent winner of the Booker Prize -quite obviously draws on Alan Palmer's work. Terms like 'mind-reading' and 'group minds' abound in the novel with lots of 'they' narration and passive forms, not to mention a scarcity of any proper names in expressing collectively shared thought.
The establishment and solidity of the genre seems to indicate some potentially contradictory observations about real and fictional minds. Above all, the way readers have received, accepted, and selected novels is a powerful argument for the real-life relevance of discursively presented fictional minds. There are elements of generic reference ("some people I know surely think just like that"), points of identification ("I feel the same, even though I couldn't find the words for it"), and subject positioning ("Wow, I want to feel this way as well"). Within this circle of imagined narrative minds, readers arguably neglect the irrelevant proposals and, at the same time, adapt new ways of thinking, feeling, and experiencing. At the other end of the spectrum, the existence of the growing literary genre enables the most writerly and 'unnatural' fictional minds. The circle of writing and reading fictional minds therefore does not indicate the reign of any traditional realism or straightforward referentiality; rather, it relies on sufficient recognition, invention, and co-construction. At the end of the story, ordinary people have a relatively shared understanding that other people have minds and mental interiority.
I have argued in this article that the categorical conception of the non-referentiality of fiction needs to be revised and replaced by a more nuanced conception of the cultural and social referentiality of fiction. This is highly relevant from the perspective of sociology as well, since the idea of non-referentiality unnecessarily sets fiction apart from social reality. By understanding fiction as an experimental lab, but without reducing its complexities, fiction can be understood as a profound theoretical project that sheds light on such concepts as identity, intersectionality, social positioning, and the mind, to mention just a few of the most obvious candidates.