Now you see it, now you don’t: methods for perceiving intersubjectivity

What work is entailed in learning skills? How do processes of enskillment enable alternative ways of seeing, feeling and acting? How can we intersubjectively meet each other when deception is regarded as prevalent? This article addresses these questions through recounting my experiences associated with learning entertainment magic. I outline how the training of competencies can involve an ongoing problematising of the relationship between seeing and knowing as well as between knowledge and ignorance. Furthermore, through recounting the affirmations and disorientations of my process of learning, I want to consider how skills acquisition can serve as a method for self-other investigation.

unfolding questions, uncertainties and reconsiderations that emerged through attending to the starting research concern with forms of work. A recurring theme will be how, again and again, my enskilling involved attempts to know others and myself. In this regard, while magic was the topic for investigation, the practice of magic itself served as a method -a method of turning toward self and other. As will be elaborated, though, this turning toward was not straightforward, not least because of the manner learning entailed both closely attending to, and coming to doubt, sensory experiences.
Drawing on traditions within hermeneutics, section 6 sets out how I came to conceptualise my exploration of self and other. In the final section I use the sensitivities cultivated to explore the relevance of this specific line of research for qualitative research more widely as it pertains to: (i) how skill learning can entail an ongoing problematising of the relationship between what counts as 'seeing ' and 'knowing'; (ii) how knowledge ignorance as well as the ability and inability to know can be constitutive of learning; and (iii) the stakes and relevancies of intersubjectivity in deceptively engaging with others.

Textual instructions 'No-clue discovery'
A spectator chooses a card and returns it to the deck. He then cuts the deck and completes the cut. His card is lost in the pack and no one -not even the magician -knows where the card is.
The magician takes the deck and begins dealing cards one at a time into the face-up heap on the table. As the magician deals, he instructs the spectator to call out the names of the cards. The spectator is asked to give no clue when his selected card shows up. He is not to pause, hesitate, blink or change his facial expression. Nevertheless, the magician claims, he will be able to detect the faintest change in the spectator's tone of voice at the exact instant the chosen card shows up.
The cards are dealt one at a time off the top of the deck. The spectator calls them out as they as are dealt. It does not matter how he calls them out; he can disguise his voice, whisper, shout or name the cards in French; when the chosen card turns up, the magician immediately announces that it is the card selected by the spectator. (Fulves, 1976: 1-2) Where to begin though? The question has particular significance in relation to learning 'entertainment magic' (or what is otherwise called 'modern conjuring') because of the comparative absence of conventional pathways for training. Whereas many other types of performance are enculturated through professionally sanctioned training as part of established organisational settings (schools, studios, etc.), both are relatively rare in the case of magic. Clubs and societies provide important collective settings for socialising skills and competencies (Jones, 2011), but their availability and make-up vary widely. In short, informal avenues for training are typical (e.g., Rissanen et al., 2014).
A phased approach to learning was adopted in this self-study in order to enable a comparison between the forms of work associated with each type of training. While, at the time of writing, I have undertaken face-to-face masterclass training and given faceto-face and virtual public shows, this article attends to forms of training undertaken in my first year. This includes learning through books, videos, mirror practice and small group performance sessions with acquaintances.
In the past, instructional books have been a common means of enabling new entrants into community traditions, while also delimiting the circulation of secreted information. As part of its extensive magic collection, Dover Publications printed eleven 'self-working magic' books by the prolific writer Karl Fulves. Based on a suggestion from the academic-magician Wally Smith, in late 2017, my pathway began with Fulves' books. Published in 1976, Self-Working Card Tricks: 72 Foolproof Card Miracles for the Amateur Magician initiated this Dover series. The first instruction is for 'No-clue discovery', and the quote above provides the opening paragraphs setting out the trick's intended effects.
How to approach the following of such instructions? While no definition of 'selfworking' is given within the book, Fulves (1976: v) describes the tricks set out as 'easy to master' because they require 'no skill'. Based on well-established themes in social research regarding the incompleteness of instructions (e.g., Livingston, 2008), however, my starting approach was that of attending to the gross and subtle forms of work needed to move from the formal, codified prescriptions to situated action.
What then was the work of following instructions of Self-Working Card Tricks? In this article, I attend to self-other dynamics. As with many tricks, envisioning was a central basis for aligning with the instructions of 'No-clue discovery'. Against my own perspective, I repeatedly imagined what others would see and hear. As I fancied at the time anyway (see section 5), this amounted to something akin to watching a video filmed from across the table. As such, my own ways of observing, reflecting and acting were taken as the analog model for attributing how others would think and act. Within philosophy of mind, the term 'simulation' refers to how we intersubjectively know the minds of others by emulating and ascribing mental states through such reasoning (Goldman, 2002).
For instance, as part of describing the mechanisms for identifying the card, 'No-clue discovery' states: Tell the spectator to replace his card on top of the packet that lies on the table. Your instructions should be something like this: 'Please place your card back in its original position in the deck.' As you speak, point with the right hand to the tabled packet. As a matter of fact, the spectator is not returning his card to its original location, but this fact is never questioned. (Fulves, 1976: 2-3) In this, the 'No-clue discovery' instructions set out a minimal sense of intersubjectivity: individuals interact and coordinate actions with one another as part of a shared activity. Many meanings given to 'intersubjectivity' across the social sciences and humanities, however, would not be applicable to this scenario. It would be questionable to characterise the instructions as involving shared (or even similar) mental states (Zlatev et al., 2008) or experiences (Brinck, 2008). In general, magic is predicated on the possibility of fostering fundamentally dissimilar states between performers and audiences. 1 This is recognised in popular portrayals of magic, which treat it as entailing deception. Magicians might proffer all sorts of verbal and non-verbal explanations for their feats, but audiences appreciate that these can function as techniques of ruse. Magicians, in turn, craft their routines in light of such anticipations.
Along these lines, when enacting the instructions in Self-Working Card Tricks, I imagined how the sequence of steps would be seen and interpreted by onlookers by trying to see myself through their scrutinising eyes. In the case of the quote above, such envisioning led me to doubt that no one would ever question the 'Please place your card back in its original position in the deck' directive. In keeping with the way skill cultivation entails both centripetal discipline to, and centrifugal derivation from, community stipulations (Downey et al., 2015), working through the instructions in Self-Working Card Tricks frequently elicited in me a concern for the kinds of actions not anticipated, and therefore the need for situated adjustments. Thus, at times, while the instructions provided the core basis for my simulations, the simulations provided the terms for assessing the adequacy of the specifics of the instructions.
Missing from the accounts of spectators by Fulves then is what seems central to the undertaking of tricks: the lived interactions between people. This is not just to say that the instructions are no substitute for hands-on experience. It is also to point out that instructions don't identify or contain all the resources needed for navigating the step-bystep undertaking of tricks. As Gallagher (2001) has contended, we rarely intersubjectively know others only through postulating or imagining mental states. Instead, interactions are typically characterised by a rich diversity of ongoing signalling that helps individuals to directly understand each other. Eye movements, facial expressions, posture, displays of emotions and the like make attempts at 'mind-reading' more like 'body-reading'. These are the kinds of expressive contingencies that I began to appreciate were not present in the instructions in Self-Working Card Tricks.

Instructional videos
While in the past instructional books served as an essential resource for many aspiring magicians, today a vast range of audiovisual instructions are available through DVDs and on-line. Videos can display a complex array of (simultaneous) bodily movements that would each require lengthy individual descriptions involving specialised terminology if codified into writing. My engagement with audiovisual instructions began in the late spring of 2018 as part of learning 'sleight of hand' manipulations through the video edition of the classic instructional book called The Royal Road to Card Magic (Hugard and Braue, 2015). Subsequently this self-training was complemented by watching video instructions of the sleights and tricks given in The Royal Road to Card Magic produced by others on YouTube.
In relation to the self-other themes of this article, I want to attend to the varied visual (re-) positioning associated with audiovisual instructions. Similar to the opening paragraphs of 'No-clue discovery', a common organisation of audiovisual aids is that they start with an enactment of the trick in question without divulging its mechanisms, and then proceed to offer step-by-step instructions. Whereas reading texts requires imagining what viewers will experience, videos enable learners to perceive and affectively experience tricks as an audience.
And yet, as with other forms of witnessing (e.g., Shapin, 1984), through my learning I came to understand audiovisual displays as far more complicated too. What training videos provide is not a demonstration that component sleights or culminating tricks can be done in general, but a demonstration that they have been executed in specific situations. The flipside of this specificity is that witnessing one enactment is no guarantee it can be executed elsewhere.
The camera angle is an obvious consideration bearing on whether an effect experienced as a viewer can be achieved as a performer. This matter is especially acute given the commonplace practice in on-line magic tutorials to offer performances solely for the eye of camera and its specific angling. I came to appreciate this dependency when practicing the trick 'Topsy-Turvy', the first in The Royal Road to Card Magic (Hugard and Braue, 2015) and one that figured in my second sleight-based card session (see below). Despite watching video after video, when practicing in front of a mirror, I just could not perform the critical turn in this trick without prominently 'flashing' a card. Only after varied attempts to adjust my hand actions over the course of a number of days did I realise that the issue was that my mirror was closer than the camera in any of the videos. I took a couple of small steps backward and the overturn seemed nearly undetectable.
Additional complications arise in making sense of what is shown through reference to what is not. For instance, multiple filming takes can be required to achieve a displayed effect, but instructors rarely acknowledge this. As a result, what an instructor can demonstrate through a one-off filming is not necessarily easy for anyone else (including the instructor) to duplicate. Indeed, learning through watching and replicating others made me more sensitive to the varied potential deficiencies of live performances. As a result, the observable perfection of any one tutorial stood as grounds for doubting that I could consistently replicate what was demonstrated.
The points in the last few paragraphs touch on wider cultural investments (Morris, 2014). Visual recordings -such as photographs and videos -are often regarded with a kind of doubleness. The common expression 'seeing is believing' signals the stock placed in visual recordings as true chronicles. And yet, as records, such visual images are also recognised as not the same thing as the events they seek to capture. By giving a particular line of sight or by foregrounding some objects, visual imaginary can mislead. Also, what is included within the image frame marks the boundaries of what has been left out -intentionally or otherwise. It is just this sort of doubleness -genuine and contrived -that tricks play upon, and indeed, require. As a result, for me instructional videos have taken on a kind of 'haunted' quality: their efforts to display point towards what is not shown; their efforts to visibly exhibit point towards the irrevocable role of trust in viewing.

Mirror practice
Practicing with a mirror is a common technique when learning magic and, as noted above, it was one that I utilised in practicing sleights. The advantages of this technique, compared to imagining how you appear, can be pronounced. For instance, undertaking sleights can be a fiddly operation, with the thumb and fingers being varyingly maneuvered. When this kind of fiddling occurs, it seems that the card manipulation should be blatant to onlookers. Practicing in front of a mirror visually confirms the considerable scope for occulted maneuvering that can be afforded, for instance, by slightly angling the deck. Although I have practiced individual sleights hundreds of times, again and again, I still feel compelled to do so in front of a mirror to confirm the card handling is not generally detectable.
In this respect then, mirror practice brings together the possibility of perceiving yourself as another person would, with the underlying knowledge and experiences of a magician. Herein, the mirror functions as a relational device along the lines it has within certain strains of conceptualising self-other relations. Jacques Lacan (1977), for instance, famously claimed specular images secure self-fantasies: against internal tumultuous bodily drives and fragmentary thoughts, one's mirror image serves as an aspirational representation of a stable, whole and coherent self. In certain respects, practicing magic entails a kind of identification with one's mirror image. Whatever one's fraught affective and physical states, as a representation of self that is observable to others, the mirror image serves as a touchstone for gauging proficiency.
Or at least in certain respects it does. My very need to revisit sleights in front of a mirror, for instance, signals the lack of definitiveness of specular self-witnessing. One limitation is that the seeing done during mirror practice is the situated seeing of one who knows what to look for. Yet, audiences may not perceive or give significance to what might be glaringly obvious to a performer. Indeed, as magicians often (but not always) strive to not bring attention to card manipulations in the first place, what might be at the centre of concern for a magician looking in the mirror may not be meant to be relevant to audiences at all (cf. Lynch, 2013). Conversely, though, while audiences might not apprehend card sleights, a common refrain in professional advice is they can be highly perceptive in sensing certain kinds of subtleties -such as when conjurers unconsciously tense up in anticipation of a difficult manipulation (e.g., Nelms, 1969 As a response to both sets of considerations, many instructions advise learners not to get preoccupied with making sleights visually perfect, but instead to engage with audiences and thereby produce a relaxed atmosphere in order to disarm their scrutiny. 2 In this way, learning through mirror practicing entails a delicate process of seeking to refine techniques through greater visual discrimination, while also attending to whether and how seeing is relevant.

Performing
By considering the kinds of work entailed in learning skills, the previous three sections recounted varied forms of practice wherein notions of others were simulated, imagined or likewise conjured up. This section recounts experiences associated with putting on card performances in order to set the stage for further reflections on the relation between enskilling, seeing and being.
As part of such performances I conducted a set of 13 recorded sessions consisting of nine tricks derived from Self-Working Card Tricks in the first half of 2018 and a set of 10 recorded sessions in the second half of the year consisting of nine 'sleight of hand' tricks largely derived from The Royal Road to Card Magic. The background and rationale for those sessions have been elaborated elsewhere (Rappert forthcoming). 3 For the purpose of the present argument, however, the sessions were run in the mode of small focus groups, combining tricks with interspersed pre-formulated questions for group discussion. The 55 different participants were largely university staff in the UK and Sweden. 4 One of the upshots I took from these performances in relation to self-other dynamics was the importance of questioning the prevalent tendency in academic and practitioner writing to theorise magic as a competitive, controlling and hierarchical activity (Rappert forthcoming). Although the encounters in the sessions entailed various forms of asymmetry regarding who held what knowledge and how directives for action were issued, noticeable for me were forms of reciprocity. For instance, as recounted in the exchange above, I needed participants to scrutinise me; without their active attention and challenging, there would not have been anything revealed/concealed. And yet, if taken too far, such regard would have made the tricks nearly impossible to pull off. In general, participants verbally (and often playfully) sanctioned each other (or themselves) when a line of action was deemed to have been taken too far (e.g., the reproach in lines 30-31 above) or not far enough (e.g., a participant was judged as not paying sufficient attention). As another area of tension, instead of understanding themselves as simply passive spectators seeking to discern how I performed the tricks, participants frequently presented themselves as deliberately engaged in co-operative relations in which they made choices about how to act vis-à-vis their situationally expectations. And yet, as suggested by lines 1-2 of the exchange above, they overwhelmingly forwarded such accounts only after being explicitly prompted by me to consider how they were contributing to the unfolding scene.
For the purposes of the present argument, however, I want to limit regard to how 'seeing' figured in the unfolding self-other interactions. Although the self-working tricks in the first 13 sessions did not require sophisticated card sleights, one of the nine did necessitate pushing a card out of the deck to glimpse it and another entailed covertly turning over a deck. On some occasions too, when the cards got out of the required order, I needed to rearrange them there and then at the table. Almost all of the nine tricks in the second set involved one or more physical sleights -false shuffles, lifting multiple cards, forcing participants to select a predetermined card, etc. My expectation was that such manipulations would be frequently detectable, not least because of my novice status. And yet, rarely did participants forward (accurate) identifications -either when asked directly by me initially or in the numerous occasions in which they offered unprompted explanations. This was so even when, as in the exchange above, individuals reported watching me 'pretty hard'. While what counts as verbally identifying a card manipulation is open to interpretation, I would put the number of such specific occasions across all of the sessions somewhere in the high single digits.
The limits of perceptions serve as one possible set of explanations for such an outcome. Fields of science have long sought to explain why sleights and other forms of trickery prove so hard to detect -including to magicians themselves (Lamont, 2006). In recent years, renewed interest under the label the 'science of magic' has emerged that utilises tricks as experimental stimuli for studying visual perception and cognitive heuristics (Kuhn et al., 2016;Kuhn, 2019). One review summed up the principles identified through this latest phase of research as: First, some things, though directly in a person's line of sight, are not perceptible at all. Second, people do not consciously perceive everything that can be perceived. Third, what is consciously perceived depends upon attention. Individuals will fail to see even what is in their direct line of sight or fail to feel an easily perceptible touch if their attention is elsewhere. Fourth, people sometimes misinterpret what they perceive. Fifth, individuals' memories fail in ways that permit changes to occur before their eyes that they do not consciously perceive. Sixth, these failures can be regularly and lawfully produced by specific manipulations of individuals' perceptual and sensory systems (Villalobos et al., 2014: 637).
In short, what is observable depends on the means of observing. Among philosophers and psychologists, much debate has taken place about what this emerging understanding of the fallibility of cognition and perception implies for the nature of consciousness and the ultimate truth status of visual perceptions (e.g., Noë, 2002).
Settling such disputes is not the agenda here. Instead I want to juxtapose these arguments about the limits of perception against the discursive work prevalent in the sessions that defined the experiential situation at hand. In his classic study, Pollner (1987: xv) identified 'mundane reasoning' as a ubiquitous form of constructing the world wherein individuals 'experience and describe themselves as "reacting to" or "reflecting" an essentially objective domain or world'. Within the traffic court proceedings he examined, for instance, witnesses to an incident could offer radically divergent accounts. Judges seeking to adjudicate 'what happened' were thus in a position of striving to determine the facts of a world taken to exist independently of how it was observed, while also necessarily being reliant on this and that situated observation. Pollner detailed how divergent accounts of witnesses got reconciled in ways that judges could both determine the 'facts of the matter' and preserve the starting presumption that there was an essentially out-there, ordinary and objective world to be found that could be taken to exist independently of knowers.
In a similar vein, participants in my sessions did not consider that what is observable depends on the means of observing. While they repeatedly acknowledged that sight can be directed away from the cards, only once was the suggestion made that perception was significantly fallible. 5 Neither were more general claims offered that what was observable significantly depended on the means of observing or reporting. In brief, participants accounted for the unfolding scene through a realistic language according to which the familiar world is out there and our senses deliver it to consciousness.
Instead of calling into question the determinacy of perception, participants to both the self-working and sleight-based sessions offered stock explanations echoing popular understandings of the mechanisms of magic -explanations overwhelmingly inaccurate to the tricks at hand or, at best, referring to highly general principles. Some of these explanations included physical sleight dexterity, cards being placed up sleeves, and hidden mirrors. Participants also frequently sought to identify tell-tale indicators in the performance that signalled the hidden mechanics. Again, the identified indicators often proved extraneous or irrelevant.
While participants sought to determine the 'facts of the matter' in a manner that preserved the possibility for an essentially 'out-there' world, as part of my self-reflections I began to notice I was doing the same. It is perhaps not surprising that in practicing (avowed) deception, a preoccupation that repeatedly came to the fore in my mind was this: might participants be deceiving me? Might they have spotted (potential) give-away signs and just decided that they should pass without comment in the name of cooperation or avoiding embarrassment? This would align with the contention of many social theorists that tact and flexibility with the truth are part of the fabric of everyday interactions (Miller, 2003). So too it would align with seasoned professional magicians' recognition of the potential (and, indeed, likelihood) for audience accommodation (e.g., Brown, 2003). Therefore I could hardly rule such deception out, whatever I heard. But I could not definitely discern it either. In short, in being prompted to look back after conducting some performances, I didn't envision what I thought I had imagined while practicing with Self-Working Card Tricks. Not only then did I come to question whether my experiences could serve as analog simulation model for others, I also came to question whether my experiences were anything like I had previously understood them. Akin to a skilful display of magic, the effect was befuddling and exhilarating in similar measure.

Strange learning
Entertainment magic often matters for the way it affectively and cognitively stirs us. 'How was that done?' is a common response to the inexplicable. Learning magic dispels much of the allure of conjuring, not least through the sheer grind of repetitive practice. And yet, as I have elaborated, while the initial wonder and mysteries of magic faded, in my learning another sense of curiosity emerged regarding what we might gloss as the 'situated accomplishment of intersubjectivity'.
One pragmatic response to the kinds of hesitation about experience outlined in previous sections is to simply set them aside in favour of focusing on 'what works'. Much of the instructional material for magic is concerned with just that: imparting learners with sufficient skills that enable them to satisfactorily undertake tricks. If we embrace rather than set aside the troubles of knowing self and other, however, what possibilities might open up?
The question of how a process of skill acquisition could serve as a means of investigating notions of self and other became central to my learning. The manner I eventually came to engage with these troubles was by theorizing conjuring through the umbrella notions of the 'familiar' and the 'strange'. In Truth and Method, Gadamer attended to ways of being that brought a sense of affirmation and belonging as well as disorientation and distance (see Kerdeman, 1998). Much of day-to-day life consists of participating in activities that are rendered familiar through our established preconceptions and conventions. Gadamer, however, was also attentive to disruptive experiences in face-to-face dialogues, reading texts, viewing art, etc. that leave us feeling 'pulled up short' (Gadamer, 1975: 270). These experiences create an awareness of how our understanding is situated between the known and the unknown. In philosophical hermeneutics, the movements of being thrown beyond and affirming what one previously understood provide the basis for (re-)forming notions of self and other. The task to be taken up is not one of covering up the tensions with being in-between the familiar and strange but bringing them out (see Gadamer, 1975: 305).
Beyond the overall descriptive parallels between philosophical hermeneutics and how I came to characterise my experiences as both affirming and disorientating, this tradition speaks to my learning because of the importance it places on how we encounter one another through the mix of the familiar and the strange. This is the topic I wish to take up in the remainder of this section. Schwandt (1999) interprets Gadamer as outlining three ways one can respond to feeling 'pulled up short'. One is for others to become an object of study in order to make predictions about their behaviour and nature. In this way, the other is rendered tractable. A second way of responding is to superficially acknowledge the importance of others' understandings, but then to impose one's own interpretation on them all the same.
In contrast, Gadamer advocated that 'understanding requires an openness to experience, a willingness to engage in a dialogue with that which challenges our self-understanding. To be in a dialogue requires that we listen to the other and simultaneously risk confusion and uncertainty both about ourselves and about the other person we seek to understand' (Schwandt, 1999: 458). That risk stems from the need to find ways of 'inducing interpretative interactions that not only expose us to the unusual and unanticipated but which also place the assumptions of our customary horizons at risk' (Davey, 2006: 4). Arguably Gadamer's thinking was not well developed though in elaborating what is needed to induce interpretative interactions such that they further openness and challenge (Kerdeman, 1998).
In engaging with Gadamer as a way into understanding the experiences recounted in previous sections, my attention turned toward considering how the self-other dialogues associated with magic can be conducted in ways that further openness and challenge. Taken in general, the manner in which attentiveness to others as well as the mix of affirmation/disorientation figure in conjuring makes it a ripe activity for reflecting on everyday assumptions. Yet, again in general, as an activity magic strikes a troubled relation toward questioning. Deception is integral to magic and recognized as such by audiences. More than this though, the anticipation for audience scrutiny itself can be used reflexively by performers to engage in forms of deception (Smith, 2015). For instance, consider lines 7-9 in the transcription in the previous section. Under the pretense explaining something about magic to the participants, I was able to undertake a sleight based on spreading the cards. As such, dialogue takes place on fraught terms. Another reason for the troubled relation toward questioning in magic is that the interactions often involve stark asymmetries in who speaks, who directs and who acts.
In terms of my efforts, I can propose ways in which I ended up promoting questioning in line with the recommendations of philosophical hermeneutics. For instance, the basic focus group-inspired organisation of the performances examined in section 5 was intended to take engagement with audiences beyond the typical affective responses that follow acts of magic (e.g., displays of surprise, curiosity, incredulity). Also, as noted in the vignettes, my ongoing reflection was attentive to anomalous experiences and the difficulties of understanding others. Through this attention I was able to take my own assumptions as a topic for consideration.
And yet, despite such affinities between the interpretative interactions suggested philosophical hermeneutics and the magic sessions discussed in the previous section, the prior analysis suggested various areas of concern regarding my engagements vis-à-vis Gadamer's call to further openness. For instance, the statements in the quote at the start of section 5 regarding being 'mugged' (lines 32-39) hint at the many fraught ethical and affective dimensions of trickery. This includes the potential for individuals to feel defensive, duped, demeaned, etc. As well, the concerns previously articulated about the indeterminate relation between seeing and knowing beg questions about what 'openness' could or should mean.

Further reflections
Having come to this mixed assessment of the extent of openness and challenge as part of the activities recounted in section 5, I wish to close this section by reflecting on the analysis given so far. As a study into self-other relations, the preceding argument has highlighted the diverse range of alternatives available for and stakes associated with performing conjuring. Such attention, however, also could be directed at the alternatives and stakes of analysing conjuring that this article has undertaken. For instance, the previous argument has outlined some of the epistemic frustrations of repeatedly trying to discern the mental states and experiences of others. Are other orientations possible?
In this regard, it is worth noting that there are many other activities involving interactional asymmetries in which those asymmetries have led to questions about how people can meet each other in ways that further, rather than reduce, dialogue. As part of this questioning, various scholars have delineated the qualities of our attention to each other. For instance, in the context of caring, Noddings (2013) juxtaposed projection and reception. Projection entails seeking to analyse and establish what another is experiencing. Reception, in contrast, entails a motivational shift. That shift calls for becoming engrossed with the other in order to attempt to feel for and become sensitive to their wants and needs -indeed to be transformed by them -even as it is recognised that it is not possible to straightforwardly access their reality. Noddings argued that reception is not about making another person into an object, because reception is not driven by a desire to make claims to knowledge. Instead, it is about receiving others. In making the case for the importance of reception in caring relations, Noddings did not seek to exclude other forms of attending. Analytical forms of projection to derive knowledge still have a place. 'What seems to be crucial' she argued 'is that we retain the ability to move back and forth and to invest the appropriate mode ' (ibid.: 35).
In looking back on my efforts to undertake and analyse my learning, I can recognise how I have been highly indebted to forms of attending based on projection -that is, attempting to establish an account of experiences; this not least for the purpose of formulating research findings. I repeatedly experienced the inability to do so as a kind of trouble or failure. It need not have been so, though. Instead, it could have been taken as the basis for the appreciation of learning as a form of being together rather than knowing the other.
In this way, through its commitments, the previous argument both provides a testament to the learning undertaken and marks the limits of learning.

Presenting conjuring
Through recounting a process of skills acquisition, this article has examined how learning can entail an inquiry into notions of self-other. In this final section, I want to conclude by comparing magic to other forms of competency learning.
A first set of conclusions relates to the place of seeing in social interactions. The mutual scaffolding between seeing and knowing has widely figured as a theme in the cultural and social analysis of skills acquisition. Roepstorff (2007), for instance, presented learning to navigate through glaciers and to read brain scans as hard won enskillments wherein refined vision underpinned adept situated practice. For O'Conner (2005), sight functioned as a taken for granted means of receiving sensory inputs that enabled her and other glassblowers to gain nuanced types of focal and subsidiary awareness.
Learning in my case certainly entailed the refinement of visual motor skills (for instance, with regard to finger positioning) through assessing actions (spreading, cutting, placing, lifting, etc. cards) against specified and inferred instructional outcomes. However, what has also come to the fore is the complex and sometimes indeterminate relation between seeing and knowing. In the practices surveyed above, seeing could not straightforwardly be taken as knowing (e.g., knowing whether physical manipulations are detectable, knowing that someone is being truthful, knowing how reliably the visual effects of sleights can be repeated). Knowing, too, fostered a questioning of what takes place in seeing, for instance in relation to what was not made visible in instructional videos and the alluring seductions of gazing into a mirror when one knows what to look for. Further, determinations of whether seeing should be regarded as relevant, and how it is made so, were part and parcel of the situational constitution of interactions.
As such, definitions of learning that depict it as a process of error detection against expected outcomes (Argyris, 1995), or of relating stimulus to responses (Lachman, 1997), or of disciplining errors to achieve greater skilfulness (Downey et al., 2015) capture only part of the dynamics surveyed in the previous sections. My learning entailed a maturing hesitancy about my claims to individual agency and control, even as I became defter in physically working cards and socially working with audiences.
As I have come to understand it then, the trick in learning magic is skilfully acting in-between certainty and uncertainty, as well as the possibilities for affirmation and not. In this way (invoking Tim Ingold's (2001) adoption of James Gibson's term), learning involved an 'education of attention'. That is to say, it involved the sensitisation of the perceptual system. However, educating attention entailed unsettling the standing of seeing, not only refining it. This unsettling took place at two levels: one, making sense of specific sensory experiences (what I was seeing as part of an encounter -looking in a mirror, watching a video, etc.) and, two, making sense of the sensory capacities in general (the possibilities for discernment given the fallibilities of perception). In this way, knowledge and ignorance were both mutually implicating and constitutive of the process of learning. As a result, I came to know, to realise I did not know, to wonder what I could know, and to doubt what I had thought I knew through my unfolding engagements. In this way too, magic as a method of inquiring did not just entail learning about self-other, but also appreciating the fraught conditions for learning.
Taking these points together with themes from previous sections, learning has entailed developing receptiveness to movement, that is, when to shift between (and set aside): -particular situated events and general descriptions; -the reliance on others' accounts and the questioning of them; -the credence given to and the distancing from specific sensory experiences; -the capacity for affirming and disaffirming sensory experiences; -the desire to know the other and the possibility of alternative forms of engagement.
By way of closing, I end with some final reflections on another set of conclusions related to how this article adds to the understanding of self-other relations and in particular the relevance of intersubjectivity. Analysing the performance of magic as an 'intersubjective' accomplishment has not implied the presence of shared meaning. Nor has it signalled the completed transformation of ignorance into practical knowledge (Goodwin, 1994: 614). Instead, intersubjectivity's relevance comes in gesturing towards a belief in the potential for mutual understanding (Susswein and Racine, 2008), a possibility that matters to the extent it is worked towards rather than ultimately secured. In simulating the instructions in a book, copying movements in a video, or watching oneself in a mirror, others were treated as 'like subjects'. This expression though implies two different senses of the term like (Benjamin, 2004: 5): both that (i) I attempted to use analog reasoning to feel through others in order to establish how they sensed the world and (ii) I recognised others were distinct individuals and thus able to have dissimilar affective states and perceptions. Both aspects were critical.
But more than this, the reconsiderations of my experience prompted calling into question my self-conceptions. In this regard, enskilling has not entailed an ever-expanding claim to physical and perceptual mastery, but rather a growing appreciation of the forms of our shared body limits and the many ways in which interpersonal interactions expose us to scrutiny by others. Gilson (2015) referred to such appreciations through the term 'intersubjective vulnerability'. She highlighted vulnerability as a fundamental condition of intersubjectivity, rather than it being a matter limited in relevancy to certain impaired, susceptible, or marginalised individuals. Through treating vulnerability in this manner, it is possible to recognise the importance of our being 'open with others, to be shaped by [others], to become a self only through relation to them; it is the condition that makes it possible for us to become who we are and will make it possible for us to become otherwise' (ibid: 231). These are the kinds of considerations that have become central for me. To be sure, asymmetries in knowledge and action between magicians and audiences commonly constitute conjuring as an activity. And yet, magicians and audiences alike make themselves vulnerable as part of the mutual dependencies of social interaction.
In light of the preceding argument then, what response can be given the question 'What kind of work is required in learning magic?' One answer is this: the cultivation of sensitivities for staying with and moving between varied orientations to our experiences with others. What follows is that attempts to analytically depict the process of learning, as in the case this article, can be evaluated, in part, through how the evidence and analysis presented enables the recognition and cultivation of diverse appreciations of our encounters with one another.