Neurorights and neurowrongs

The easiest way in to thinking about MI is to consider potential threats to it. We might term such threats neurowrongs. Bublitz and Merkel5 imagine some striking possibilities, such as retailers using Flash movies for subliminal priming of customers and psychiatrists covertly using transcranial magnetic stimulation (TMS) in order to increase patients’ truth-telling. Ienca and Andorno14 imagine even more extreme scenarios, such as unauthorised interference with brain-computer interface (BCI) devices (eg, limiting or controlling function of BCI-operated wheelchairs, artificial limbs, or visual prostheses; see also Ienca and Haselager23), selective boosting or erasure of memories by modulating synaptic connections (on which they cite, specifically, the work of Nabavi et al 24 on techniques for achieving these effects), and potential applications of TMS in warfare to enhance and/or suppress brain processes. They draw this latter application of TMS from a 2009 National Research Council report25 (pp. 84–85) (and see also recent reports of animal testing aiming to recapitulate symptoms of Havana syndrome in animals using pulsed radio frequency waves).26 While these scenarios are speculative, they illustrate the kinds of extreme possibilities that future advances in neuromodulation might occasion. In all of these cases, the victim’s mind is impinged upon in an objectionable way.

Gilbert et al 27 provide a final case that is especially instructive due to having actually occurred. They describe a patient suffering from severe chronic epilepsy whose seizure-detection implant was explanted against her wishes when the manufacturer (Neurovista) went out of business. The result was significant disruption to her autonomous agency, which had become ‘symbiotic’ with the device, as well as ‘a sense of confusion, disorientation, and anxiety as she navigates a new normal without the technology that had become an integral part of her cognitive and emotional processes’ (p. 785). Here the causal structure of the case is different: a beneficial intervention was removed, rather than a deleterious intervention being introduced. All the same, the case as described by Gilbert et al strikes me as quite a serious violation of MI.

To make good on that thought, however, we will need to clarify what exactly MI and the right to it are supposed to be. We can begin this task by considering the dominant autonomy-based conception of MI.

The ABCs of MI

According to most theorists, MI is principally about autonomy14 28–32 or something in the near vicinity, such as a self-determination,5 33 34 authentic choice,35 or self-ownership and personal sovereignty.36 This autonomy-based conception (ABC) admits of some variation that is worth unpacking.

Some theorists suggest that (1) MI is a person’s control or authority over her own mind and its associated brain processes, as against unwanted interference. Control is here understood as control over the contents of one’s own mind, such that the contravention of our autonomous preferences about these contents violates MI. Perhaps anticipating charges of overbreadth, other theorists further specify the form of control at issue. This variant of the ABC says instead that (2) MI is a person’s control over access to her mind and underlying neural processes, as against unwanted interference. Here, the autonomous preferences relevant to MI are restricted to those about the sources of influence, rather than the particular contents that they cause. It is contravention of our autonomous preferences about which sources will influence us (and how) that constitutes a violation of MI.

Other theorists, by contrast, think of MI as a state rather than a form of control. On this style of view, MI is just the absence of interference itself, such that: (3) MI is a person’s undisturbed state of not being mentally interfered with. Such views are still a variant of the ABC, however, for this reason: for these authors, what counts as interference is determined by an individual’s exercise (or lack thereof) of her autonomy. MI is the state of not having our autonomous preferences regarding our minds contravened. Which autonomous preferences count, of course, can be specified as either of the control variants (1) and (2) specify them.

In consequence, there is not much distance between control and state variants of the ABC. One camp identifies MI with the autonomous control whose non-violation consists in the undisturbed state that the other camp identifies with MI. Both camps agree that autonomy is central to MI and in analysing a corresponding right to MI as a negative right of non-interference. This right then either coincides with MI itself (on state conceptions of it) or safeguards the conditions of its exercise (on control conceptions). MI and the right to it are violated when our minds are interfered with contrary to our autonomous preferences. To make all of this clearer, we can examine some specific accounts of MI offered in the extant literature.

One influential account, Ienca and Andorno’s, emphasises the importance of protection against ‘unauthorised alteration’ of our mental life (14, p. 18), where the significance of the alteration being unauthorised is that this renders it contra-autonomous. Their more formal definition goes as follows:

For an action X, to qualify as a threat to mental integrity, it has to: (i) involve the direct access to and manipulation of neural signaling (ii) be unauthorized – i.e. must occur in absence of the informed consent of the signal generator, (iii) result in physical and/or psychological harm (Ibid.).

Condition (ii), in particular, makes clear the conceptual connection to autonomy. Because Ienca and Andorno’s definition is about what constitutes a threat to MI, it is ambiguous between the control and state variants of the ABC. It is most naturally interpreted as a state variant of the ABC, but could also be put in terms of control. Ienca and Andorno’s variant of the ABC is fairly stringent. It involves not only a lack of consent, but direct invention in neural processes and the presence of harm.

Subsequent accounts, however, have tended to broaden their conceptions of MI (and the right to it). Lavazza,31 for example, frames his own analysis against the backdrop of Ienca and Andorno’s. Lavazza loosens their (i) and drops (iii) altogether to arrive at the following definition:

Mental Integrity is the individual’s mastery of his mental states and his brain data so that, without his consent, no one can read, spread, or alter such states and data in order to condition the individual in any way (31, p. 4).

Here we have a clear instance of the control variant of the ABC, since MI is said to be an individual’s ‘mastery’ of her own mind and information about it.

So, while there is some intramural disagreement between ABC-theorists, they are unified by the idea that MI and the right to it are grounded in autonomy. More specifically, the ABC understands it primarily as a negative right of non-interference (a right against others affecting a person in certain ways, rather than a positive right to some good). It is fair, I think, to describe the ABC as an emerging consensus in the nascent literature on MI. But it faces difficulties that suggest something has gone awry. We can turn now to those difficulties, which will lead us to seek an alternative.

Revising the ABC

An important worry for the ABC is that it makes MI, and the concomitant right to it, far too broad. There are two distinct forms of overbreadth worth considering. First, there is overbreadth that results from understanding MI as threatened by both inputs and outputs. Our discussion so far has revolved around inputs: unacceptable interventions in a person’s mental life, causing it to be some way it should not be. But Blumenthal-Barby and Ubel37 think that the field also mistakenly counts a second very different phenomenon as threatening MI: outputs, the collection and transmission of information about a person’s mind and brain. Protecting both inputs and outputs under the rubric of a single right, Blumenthal-Barby and Ubel argue, leads to an implausibly broad and theoretically disunified conception of MI.

The charge is, I think, apt in the case of Lavazza’s31 account and others like it that understand MI as control over both one’s mind and information about it (or about underlying brain processes). Blumenthal-Barby and Ubel seem quite right that inputs and outputs are too dissimilar to be covered by a single right. Outputs chiefly concern mental privacy, which is distinguished from MI by most theorists of neurorights. This distinction is appropriate because a threat to integrity seems, conceptually, to involve a threat of deleterious alteration. But mental outputs can, at least in principle, be passively collected without at all disturbing the individual’s ongoing mental processes. While clearly worrisome, mere collection of information about the mind has a fundamentally different character than actively intervening in the mind. Theorists of MI, then, can reasonably respond to Blumenthal-Barby and Ubel’s worry by understanding inputs alone as fundamentally relevant to MI and holding that outputs require their own separate analysis in terms of concepts like mental privacy. This is compatible with recognising, of course, that outputs can make for more effective, individually tailored inputs. Outputs might make violations easier or more profound, but no amount of outputs alone violate MI (even if they do, plausibly, violate some other right).

A second, more serious form of overbreadth results from the fact that mental interference is ubiquitous.36 37 Ordinary life confronts us with a steady stream of mental inputs via sense-perception. Our minds are ‘interfered’ with by such commonplace occurrences as hearing a plane fly overhead, being confronted with street signs and traffic lights, and being engaged in conversation by other people. Anything that affects the content of our experience can be said on that basis to influence our minds, and any such influence on our minds is at least a potential influence on our behaviour as well. To say that MI is violated even by these sorts of unremarkable occurrences would be to trivialise it, and to make any supposed right to it terribly implausible. This brings out a serious inadequacy in the ABC. Any conception of MI must draw a distinction between mere mental influence and unacceptable mental interference, neurotechnological or otherwise. But this distinction has not been well articulated.

We can make progress on that score by going back to the beginning of MI discourse. The first to explicitly theorise about MI by that name was A. T. Welford.38 Welford’s concern is not neurotechnological interference in the mind, since neuromodulation research was during his time still in its infancy. Instead, we find his interest to be the then-novel conception of death as brain death. This conception of death, Welford thinks, implies a corresponding conception of life. And this conception of life he calls ‘mental integrity’. In explicating what he means, he offers a parallel between bodily and mental integrity. Both, Welford thinks, should be understood as forms of functional integrity. Here is how he puts it:

Death is normally conceived as having occurred when there is an irreversible cessation of bodily function, such that disintegration of the body will follow. Recently, especially in France, it has been argued that the individual as a whole is dead when his brain is dead—that is, when the seat of his personality and of the coordination of his thought and action which make him a human being has irreversibly ceased to function. Both these definitions are essentially in functional terms, so that we arrive at a concept of life as a state of functional integrity (38, p. 1136).

For Welford, an individual is a whole whose existence depends on the integrated function of its parts. For the body, integrated function is biological life, and its irreversible loss biological death. Something analogous is true of the mind. The life of a person, as distinct from a merely biological organism, is integrated function (‘coordination’) of various mental capacities (for Welford, thought, action, emotion, and memory). The death of a person is the irreversible loss of this coordination, and so of personality.

Unfortunately, Welford has little else to say about the specific capacities that underlie MI, or what it means for mental capacities to be functionally integrated. But this idea of MI as functional integration is also scattered throughout the more recent MI literature, often without explicit engagement with Welford. Bublitz33, for example, gets at something similar in considering the nature of integrity rights. ‘[W]hat is an integrity right?’ he asks. ‘Although there is no settled understanding, a right to the integrity of X seems to denote the preservation of the intactness, unity or identity of X. Alterations of these features then constitute interferences’ (§30.4.2). While Bublitz thinks that the best substantive conception of MI should be filled out in terms of self-determination, he recognises that the basic concept of MI itself has something to do with mental integration.

Similarly, Fuselli32 understands MI partly in terms of autonomy, but appeals as well to language like ‘mental balance’ (p. 419), ‘psychical borders’ (p. 424) and ‘the psychophysical unity of the subject’ (p. 426). He also relates MI specifically to integration:

[O]nly what is integral – what is not fragmented, not divided, not disjointed, not dispersed – can be properly called individual. Only what is not disjointed because of its individuality can have its own identity and autonomy in relating to itself and to somebody else or something else (p. 423).

Like Welford, Fuselli understands individuality as dependent on integration. Valera39 appeals to similar language (eg, ‘psychosomatic unity’). In dialogue with Fuselli, he also says, ‘Individuality needs integrity: a subject is an individual (from the Latin individuus, i.e., indivisible) only if its mental ecology is intact’39 (p. 102). Hildt40 likewise proposes an understanding of MI as ‘being mentally whole, unimpaired, or undivided’ (p. 91). Keeling and Burr35 propose an allostatic view of MI according to which it involves both an internal regulatory equilibrium and a similar equilibrium between the mind and the environment. And Fins’41 emphasis on positive as opposed to merely negative neurorights suggests a similar way of thinking about MI, one that anticipates some of the ideas of my own proposal.

These remarks all point in a common direction, towards a conception of MI as functional integration of the mind. Though all of these contemporary theorists retain autonomy-centric views of MI, they have rediscovered Welford’s thought that MI consists in something more fundamental that is a precondition of autonomy and other high-level features of a person. Valera39 sums the idea up nicely: MI is ‘the condition of possibility for other human dimensions, such as freedom, autonomy, and agency’ (p. 102).

This preconditional picture of MI suggests a way to revise the ABC to avoid overbreadth. Rather than understanding violations of MI in terms of the frustration of particular autonomous wishes about the contents of our minds, or in terms of particular instances of consent or its absence, we can shift to a different level of analysis.2 This revised ABC (‘ABC-R’) concerns itself exclusively with effects on our underlying general capacity to form and enact autonomous preferences and to consent. It is interference at this level, I propose, that threatens MI in particular. What is most distinctive about the kinds of cases we considered at the outset is that the victims’ standing abilities to evaluate, deliberate, choose, and act are themselves interfered with, such that their usual capacity for autonomy is (at least temporarily) diminished. The very autonomous self from which preferences and consent issue in the first place here comes under threat.

Zohny et al 42 independently arrive at a similar, but narrower, view. On their account, MI is violated by mental interventions that bypass rational control and cause alienated mental states. This is a helpful proposal, but one that seems to me to narrow the scope of MI too significantly. Bypassing a person’s rational control is one way of diminishing her autonomous agency, but not the only way. A second way is to engage her autonomous agency, but with inputs that will erode it over time. Here addiction may be an instructive example. A third way is overwhelming a person’s autonomous agency with inputs too complex to adequately navigate. Here an instructive (though not entirely uncontentious) example may be contemporary online consumers confronted with a dizzying array of choices, price points, reviews of varying reliability, and real-time targeted advertisements.

On the ABC-R, a right to MI is much more plausible, since it now appears to be a special case of a more general right safeguarding the basic conditions of our autonomy. Such an account also has the benefit of bringing theorising about MI into alignment with other areas of contemporary technology ethics. Similar thinking about autonomy has been deployed, for example, in arguments for the practice of digital minimalism.43–45

Accounting for the non-autonomous

We might think, then, that the ABC-R gives us a satisfactory theory of MI. Unfortunately, the ABC-R still faces a different kind of problem: making sense of the MI of the non-autonomous. In the case of bodily rights, we properly draw a distinction between bodily autonomy and bodily integrity, where the latter is something like ‘the right to have one’s own body whole and intact’46 (see also Herring and Wall’s analysis).47 This right is appropriately accorded even to those who lack autonomy, such as children and those with profound intellectual disabilities. These individuals, too, have a right to the preservation of their bodily wholeness and intactness even in the absence of an autonomous preference to that effect. These individuals’ rights vis-à-vis their bodies will need to be grounded, then, in something other than autonomy. In like fashion, these non-autonomous individuals plausibly have rights against certain forms of mental violation, too. We would presumably not tolerate arbitrary interference with their mental processes despite their lacking any autonomous preference against such interference. But such rights cannot, of course, be autonomy rights. MI must then be grounded in something more general and fundamental than autonomy—something that is accompanied by or involves autonomy in the autonomous, but is present also in the non-autonomous.3

To get a handle on what that might be, consider again the preconditional idea of the last section. A view of that kind need not appeal to a single, highly sophisticated capacity (autonomy) in articulating the normative significance of MI. We can instead understand MI in terms of a plurality of simpler capacities, ones that can be combined in sophisticated ways to yield higher-order capacities such as autonomous agency. But the capacities need not rise to that high level in order to be worthy of protection. They are worthy of protection in themselves, including as they manifest in the non-autonomous. Or so I will now suggest.

The simpler capacities I propose are the capacities for (1) affect, (2) cognition, and (3) volition. These are present not only in autonomous adults, but also in children and adults with profound intellectual disabilities, as well as a wide variety of non-human animals. Even in rudimentary form, these capacities are deserving of protection because each is the basis of certain fundamental interests. The value of MI is that it is the precondition of possessing these interests. On my account, the right to MI is a right against diminishment of these interest-grounding mental powers.

Integrity and interests

By an individual’s interests, I mean the fundamental respects in which her life has first-personal significance. We have interests insofar as our lives matters from the inside, in ways dependent upon and only fully graspable through our own perspective (see also Willaschek,48 who arrives at a similar conception of value via a rather different route). Such interests come, it seems to me, in three basic varieties: (1) affective interests deriving from our nature as feeling subjects, (2) epistemic interests deriving from our nature as inquiring subjects, and (3) agential interests deriving from our nature as willing subjects.

Denizens of the philosophical literature on welfare will recognise these as the three kinds of capacities picked out by values-based theories of welfare.49–53 Contra such theories, I believe it is instead most plausible to identify our welfare solely with interests of the first kind, such that epistemic and volitional interests are personal but non-welfare interests. The account of MI on offer, though, does not stand or fall with that claim. If the claim is incorrect, and welfare does turn out to be the joint product of affective, epistemic, and volitional elements, a welfare-based account of MI would instead be plausible. Even then, however, it would still make good sense to distinguish the three kinds of interests, despite bringing them under the common heading of welfare. With that in mind, we can now consider these three kinds in detail.