The Experience of Trying

Josh Shepherd– Junior Research Fellow in Philosophy, Jesus College; Postdoctoral Research Fellow, Oxford Centre for Neuroethics; James Martin Fellow, Oxford Martin School

What kinds of con­scious exper­i­ences accom­pany (and per­haps assist) the exer­cise of con­trol over bod­ily and men­tal action?

For answers on this and related ques­tions, one might turn to the rap­idly grow­ing lit­er­at­ure on the so-called ‘sense of agency.’ The sense of agency is sup­posed to be some­thing exper­i­en­tial and related to action, but I think it is fair to say that there is little unity in the ways sci­ent­ists deploy the term. Andreas Kalckert (2014) writes that the sense of agency is ‘the exper­i­ence of being able to vol­un­tar­ily con­trol limb move­ment.’ Hauser et al. (2011) write that the sense of agency is ‘the exper­i­ence of con­trolling one’s own actions and their con­sequences.’ Damen et al. (2015) write that ‘The sense of agency refers to the abil­ity to recog­nize one­self as the con­trol­ler of one’s own actions and to dis­tin­guish these from actions caused or con­trolled by oth­er sources.’ Chambon et al. (2013) write ‘sense of agency’ refers to the feel­ing of con­trolling an extern­al event through one’s own action.’ David et al. (2008) write ‘The sense of agency is a cent­ral aspect of human self-consciousness and refers to the exper­i­ence of one­self as the agent of one’s own actions.’ These glosses vari­ously emphas­ize vol­un­tary con­trol of limb move­ment, con­trolling actions and con­sequences, con­trolling con­sequences through action, abil­it­ies of recog­ni­tion and dis­crim­in­a­tion, and an exper­i­ence of one­self as agent. While these glosses might share a neigh­bor­hood, they dif­fer in details that are, argu­ably, quite import­ant if one wants to under­stand the kinds of exper­i­ence at issue in bod­ily (and men­tal) action control.

In my own work, then, I have eschewed use of the term sense of agency, pre­fer­ring instead to start with a more detailed account of the phe­nomen­o­logy. Consider, for example, what I have called the exper­i­ence of trying.

To get a grip on this kind of exper­i­ence, con­sider lift­ing a heavy weight with one’s arm. Doing so, one will often exper­i­ence ten­sion in the elbow, strain or effort in the muscles, heav­i­ness or pull on the wrist, and so on. In addi­tion, there is an aspect of this exper­i­ence that is not to be iden­ti­fied with any of these haptic ele­ments, or with any con­junc­tion of them. When lift­ing the heavy weight, one has an exper­i­ence of try­ing to do so. Put gen­er­ally, then, we might say that the exper­i­ence of try­ing is an exper­i­ence as of dir­ect­ing activ­ity towards the sat­is­fac­tion of an inten­tion (this is not to say that pos­sess­ing a concept of inten­tion or of an intention’s sat­is­fac­tion is neces­sary for the capa­city to have such exper­i­ences). In the example at hand, it is a phe­nom­en­al char­ac­ter as of dir­ect­ing the move­ments of the arm.

With this much, many appear to agree. David Hume speaks of the ‘intern­al impres­sion’ of ‘know­ingly giv­ing rise to’ some motion of the body or per­cep­tion of the mind. His lan­guage sug­gests that he regards the ‘giv­ing rise to’ as fun­da­ment­ally directive.

It may be said, that we are every moment con­scious of intern­al power; while we feel, that, by the simple com­mand of our will, we can move the organs of our body, or dir­ect the fac­ulties of our mind. An act of voli­tion pro­duces motion in our limbs, or raises a new idea in our ima­gin­a­tion. This influ­ence of the will we know by con­scious­ness. (2000, 52)

 Further evid­ence for this point is that Hume thought of this experience-type as whatever is shared in both suc­cess­ful and failed actions:

A man, sud­denly struck with palsy in the leg or arm, or who had newly lost those mem­bers, fre­quently endeav­ours, at first to move them, and employ them in their usu­al offices. Here he is as much con­scious of power to com­mand such limbs, as a man in per­fect health is con­scious of power to actu­ate any mem­ber which remains in its nat­ur­al state and con­di­tion. (53)

More recently Carl Ginet has asser­ted a very sim­il­ar view.

It could seem to me that I vol­un­tar­ily exert a force for­ward with my arm without at the same time its seem­ing to me that I feel the exer­tion hap­pen­ing: the arm feels kin­es­thet­ic­ally anes­thet­ized. (Sometimes, after an injec­tion of anes­thet­ic at the dentist’s office, my tongue seems to me thus kin­es­thet­ic­ally dead as I vol­un­tar­ily exer­cise it: I then have an illu­sion that my will fails to engage my tongue.) (1990, 28)

Are these philo­soph­ers right? In a recent paper (Shepherd 2015) I argue for a pos­i­tion that seems (to me) to indic­ate the answer is yes. This is the view:

Constitutive view. The neur­al activ­ity that real­izes an exper­i­ence of try­ing is just a part of the neur­al activ­ity that dir­ects real-time action control.

 My argu­ment – very briefly – is this. There is no good empir­ic­al reas­on to deny this view. And there is some empir­ic­al reas­on to adopt it. In what fol­lows I’ll offer a shortened ver­sion of the first part of this argument.

Why do I say there is no good empir­ic­al reas­on to deny the view? The best empir­ic­al reas­on would stem from applic­a­tion of a cer­tain kind of the­ory of the ‘sense of agency’ to exper­i­ences of try­ing. This the­ory seeks to estab­lish that some ver­sion of a com­par­at­or mod­el of the sense of agency is cor­rect. According to the com­par­at­or model:

an inten­tion pro­duces overt action by inter­act­ing with a tangled series of mod­el­ing mech­an­isms that take the intention’s rel­at­ively abstract spe­cific­a­tion of a goal-state and trans­form it into vari­ous fine-grained, func­tion­ally spe­cif­ic com­mands and pre­dic­tions. An inverse mod­el (or ‘con­trol­ler’) takes the goal state as input and out­puts a motor com­mand designed to drive the agent towards the goal-state. A for­ward mod­el receives a copy of the motor com­mand as input and out­puts a pre­dic­tion con­cern­ing its likely sens­ory con­sequences. Throughout action pro­duc­tion, the inverse mod­el receives updates from vari­ous com­par­at­or mech­an­isms. On stand­ard expos­i­tions of the mod­el (e.g., Synofzik et al. 2008), three types of com­par­at­or mech­an­ism are pos­ited. One com­pares the goal-state with feed­back from the envir­on­ment, and informs the inverse mod­el of any errors; a second com­pares the goal-state with the for­ward model’s pre­dic­tions, and informs the inverse mod­el of any errors; a third com­pares the for­ward model’s pre­dic­tion with feed­back from the envir­on­ment, and informs the for­ward mod­el (so as to devel­op a more accur­ate for­ward mod­el). (Shepherd 2015, 5)

 On a com­par­at­or account of agen­t­ive exper­i­ence, when pre­dicted and desired (or, at slower time scales, pre­dicted and actu­al) states match, the giv­en com­par­at­or ‘codes’ the activ­ity as self-generated. This code is then sent to a sys­tem hypo­thes­ized to use it in gen­er­at­ing the sense of agency. Proponents of the com­par­at­or account recog­nize that this is not a com­plete explan­a­tion of agen­t­ive exper­i­ence, but they main­tain that this match­ing pro­cess “lies at the heart of the phe­nomen­on” (Bayne 2011, 357).

Notice that accord­ing to the com­par­at­or account, the neur­al activ­it­ies that real­ize agen­t­ive exper­i­ence are not dir­ectly involved with action gen­er­a­tion and con­trol. If a com­par­at­or mod­el can account for the exper­i­ence of try­ing, then the con­stitutive view is likely false. Of course, this account was not designed to explain exper­i­ences of try­ing, but rather the sense of agency. Can a com­par­at­or account be exten­ded to exper­i­ences of trying?

I argue from self-paralysis stud­ies that the answer is no. In these stud­ies, exper­i­menters para­lyzed them­selves with neur­omus­cu­lar blocks that left them con­scious, and then attemp­ted to per­form vari­ous actions. Regarding the res­ult­ant exper­i­ences, here is what Simon Gandevia and col­leagues report.

All repor­ted strong sen­sa­tions of effort accom­pa­ny­ing attemp­ted move­ment of the limb, as if try­ing to move an object of immense weight. Subjective dif­fi­culty in sus­tain­ing a steady level of effort for more than a few seconds was exper­i­enced, partly because there was no visu­al or aud­it­ory feed­back that the effort was appro­pri­ate, and because all sub­jects exper­i­enced unex­pec­ted illu­sions of move­ment. As examples, attemp­ted flex­ion of the fin­gers pro­duced a feel­ing of slight but dis­tinct exten­sion which sub­sided in spite of con­tin­ued effort, and attemp­ted dor­siflex­ion of the ankle led to the sen­sa­tion of slow plantar flex­ion. Further increases in effort repeatedly caused the same illus­ory move­ments. (Gandevia et al. 1993, 97)

As I note in (Shepherd 2015):

[P]articipants had exper­i­ences of try­ing to move a fin­ger or ankle in a cer­tain dir­ec­tion. And par­ti­cipants had exper­i­ences of the rel­ev­ant fin­ger or ankle mov­ing in the oth­er dir­ec­tion. This indic­ates that the exper­i­ence of try­ing is both caus­ally linked with and dis­tinct from the exper­i­ence of the body mov­ing. (7)

This also looks like con­firm­a­tion of the claims made by Hume and Ginet, and indic­a­tion that a com­par­at­or mod­el does not work for exper­i­ences of try­ing. Nothing like a match­ing pro­cess appears to under­lie these experiences.

This leaves open a num­ber of inter­est­ing ques­tions. Do we have pos­it­ive empir­ic­al reas­on to adopt the con­stitutive view? How do exper­i­ences of try­ing relate to oth­er agen­t­ive exper­i­ences – exper­i­ences of action, per­cep­tu­al exper­i­ences in action, exper­i­ences of con­trol or of error in action, and so on? I deal with some of these ques­tions in my (2015). I deal with oth­ers in work in pro­gress. Dealing with all of them is more than enough work for much more than one person.



Bayne, T. (2011). The sense of agency. In F. Macpherson (ed.), The Senses. Oxford: Oxford University Press: 355–374.

Chambon, V., Wenke, D., Fleming, S. M., Prinz, W., & Haggard, P. (2013). An online neur­al sub­strate for sense of agency. Cerebral Cortex, 23(5), 1031–1037.

Damen, T. G., Müller, B. C., van Baaren, R. B., & Dijksterhuis, A. (2015). Re-Examining the Agentic Shift: The Sense of Agency Influences the Effectiveness of (Self) Persuasion. PloS one, 10(6), e0128635.

Ginet, C. 1990. On Action. Cambridge University Press.

Hauser, M., Moore, J. W., de Millas, W., Gallinat, J., Heinz, A., Haggard, P., & Voss, M. (2011). Sense of agency is altered in patients with a putat­ive psychot­ic pro­drome. Schizophrenia research, 126(1), 20–27.

Hume, D. (2000). An Enquiry Concerning Human Understanding: A Critical Edition, (ed.) T. L. Beauchamp. Oxford University Press.

Kalckert, A. (2014). Moving a rub­ber hand: the sense of own­er­ship and agency in bod­ily self-recognition.

Shepherd, J. (2015). Conscious action/Zombie action. Noûs.

Synofzik, M., Vosgerau, G. and Newen, A. (2008). Beyond the com­par­at­or mod­el: A mul­ti­factori­al two-step account of agency. Consciousness and Cognition 17(1): 219–239.