Ability’s Two Dimensions of Robustness, Sophie Kikkert (LSE)

Featured Image

This is a post by Sophie Kikkert (LSE), a speaker from our Spring 2021 Seminar Series. If you would like to attend a seminar, please sign up using this link. You can also download the full schedule here.


When we attribute to someone the ability to use MS Excel, speak Turkish, or change a tire, we take it that they can successfully perform these acts ‘across counterfactual scenarios’. With this I mean that often, when we say someone has an ability, this indicates that it isn’t merely possible for them to perform some act as things actually stand, but across some range of other scenarios too. For instance, someone with the ability to replace a tire might be expected to succeed at this task both today, while on the road, and tomorrow, when parked at home, as long as some set of background requirements is met (e.g. she has access to the right tools and is awake). The notion of counterfactual stability, or what I call ‘robustness’, is an important feature of philosophical accounts of ability. The fact that abilities can be more or less robust renders them gradable, and provides us with at least one way (but as I will argue, really several ways) in which we can compare people’s abilities. It also explains how, when we practice our abilities, we often aim to improve their robustness, their reliability across certain circumstances.

Which counterfactual scenarios are relevant when spelling out what it takes to have an ability is a matter of debate. No fully general answer can be provided, since the kind of scenarios that matter will differ based on the context and ability we’re interested in. After all, the circumstances in which we’d expect someone with the ability to swim to succeed are quite different from those in which we’d expect a dentist with the ability to draw a tooth to succeed. However, I think we can say something general about the structure of the set of relevant scenarios. Specifically, it seems to me that abilities can be robust along two different dimensions. The following example should help make this intuitive:

Surgeons A and B are experts in aortic valve replacements (AVRs). While A has a higher success-rate overall, B is at less risk of failing to successfully complete the procedure when operating on patient X.

Both A and B have, in some sense, the ability to perform AVRs, but the pattern of scenarios in which they would succeed differs. A is likely to succeed across a wide range of circumstances in which she’s asked to perform the operation. Let’s call that Robustness Type I. However, in the case of patient X, A need not be in a good position to exercise her ability. She might fail – or might have quite easily failed – to perform that specific operation successfully. Things are different for B. Although B is not as reliable across a wide variety of circumstances – for instance, she might only succeed when operating on men under the age of 60 -, in the case of patient X, B exercises her ability, which guarantees success, and ensures she could not easily have failed. Even if things had been a little bit different, she’d still have performed a successful AVR. Let’s call this Robustness Type II.

Distinguishing these two dimensions helps synthesise two alternative views concerning the sense in which abilities are robust. On the one hand, some have argued that having an ability requires that, as things stand, there is some act the agent can perform which guarantees success1,2. On the other, some propose that the robustness involved in ability has to do with the variety of situations in which it is open to the agent to perform some act3,4. These two views are sometimes thought of as conflicting. Proponents of the second view respond to the first that ability is hardly a guarantee of success: Also able agents sometimes fail. But, an advocate of the first view might worry, the fact that someone acts successfully across a wide range of scenarios isn’t sufficient to rule out luck, which is required when we attribute abilities: It may well be possible for me to roll a number lower than six with a fair die across a whole range of scenarios, but that seems to be due to good odds rather than ability. However, if there are indeed two dimensions along which abilities can be robust, that explains how each view gets something right. It is true that even able agents sometimes fail, because Robustness Type I doesn’t guarantee of success. It rather captures a high success-rate. It is also true that in a specific scenario, for an act to count as an exercise of ability, its success can’t be lucky. We must somehow distinguish between success due to ability and lucky success. This is what Robustness Type II can help us with.

How we understand the kinds of robustness involved in ability matters, because when we aim to find out about people’s abilities (for example through tests), it is important to clarify what kind of pattern we’re looking for. Of course, as explained earlier, which exact scenarios are relevant will depend on the ability and context in question. And when specifying this, we’ll have to think carefully about how we want to distinguish cases where the background conditions aren’t right from cases where the agent lacks the ability we’re after. However, having a better understanding of the structure of the robustness required will be a good start. Besides its import in the practical realm, the structure of different kinds of ability attributions is also important because, within philosophy, ‘ability’ is regularly called upon to help account for other phenomena5, such as knowledge6, qualia7, and scientific understanding8. Such accounts often draw on ability to confer some sort of stability or reliability upon the phenomenon accounted for. However, for any account drawing on ability to explain something else to be persuasive, it will be important to specify precisely which sense of ability – involving which kind of robustness – one has in mind. If the notion of ability is to be more than an ‘unexplained explainer’, clarifying the sense in which the ability one is referring to is robust is essential.


About the Author

Author PhotoSophie is a PhD student and keen class teacher at the London School of Economics. She is interested in what it means to have and exercise abilities, and in how abilities relate to the options and opportunities people have. Besides developing a formal framework that helps us better understand what abilities are and how they behave (partly presented in this BPPA talk), her most recent research asks how tests help us find out about people’s abilities, and what role context-sensitive and socially constructed norms play in determining whether someone has an ability or not. In future work, she would like to apply her findings within the broad domains of (team) work, education, sports and disability studies. You can contact Sophie at s.m.kikkert@lse.ac.uk.



  1. Brown, M. (1988). On the logic of ability. Journal of philosophical logic, 17(1), 1-26.
  2. Horty, J. F. and Belnap, N. (1995). The deliberative stit: A study of action, omission, ability and obligation. Journal of philosophical logic, 24(6), 583-644.
  3. Maier, J. (2018). Ability, modality and genericity. Philosophical studies, 175(2), 411-428.
  4. Mandelkern, D., Schultheis, G., and Boylan, M. (2017). Agentive modals. Philosophical review, 126(3), 301-343.
  5. Vetter, B. (2019). Are abilities dispositions? Synthese, 196, 201-220.
  6. Greco, J. (2009). Knowledge and success from ability, Philosophical Studies, 142(1), 17-26.
  7. Lewis, D. (1990). What experience teaches. In W.G. Lycan (ed.), Mind and cognition. Blackwell.
  8. De Regt, H. (2017). Understanding scientific understanding. Oxford University Press.

Leave a Reply

Your email address will not be published. Required fields are marked *

Scroll Up