By Jemimah Thompson
Entrenched within discussions of the nature of autonomy is a practical concern for how we are to respond, morally and politically, to agents’ preferences. Liberal tradition dictates that an agent’s preferences are to be respected—recognised as valid and pursued uninhibited, so long as others’ preferences are likewise uninhibited—purely by virtue of being the agent’s own. [1] The protection of this assumption is motivated by its integral place in a respect for persons, and the worry that requiring the worth of agents’ preferences to be otherwise established would undercut agents’ authority over their lives and thus deny the presumed value of their personhood. The very grounds of this assumption, however, have been challenged by inferences regarding the social contingency of preferences. The origin of even the most (apparently) freely-held preference cannot be so securely identified in the sovereign agent as once supposed. Various perspectives, including philosophical and psychological explorations of preference-forming processes [2], economic theories of preference endogeneity [3], and feminist insights into internalised oppression [4], have all arrived at this conclusion. While these collectively work to question the basis upon which a respect for persons ought to be established, it is the feminist perspectives that demand the most stringent rethinking of preferences as adaptive. Preferences having content such that an agent appears to endorse her own oppression, and so be complicit in sustaining oppressive conditions, pose a significant challenge to the assumption that equates respect for persons with unquestioned acceptance, and promotion, of an agent’s every preference.
Concurrent to the preoccupation with liberal respect for preferences is a desire to maintain value pluralism as a political principle. As respect for preferences is positioned in opposition to paternalistic political interventions, value pluralism is at odds with a political project that adheres to an account of the good—that is, a universalisable condition or set of conditions for human flourishing. It is thus seemingly in the political interest of a theory of adaptive preference to avoid an appeal to a determinate account of the good, and thereby avoid both the difficulty of developing an agreed account and the worries that accompany ensuing paternalistic intervention. An attempt to explain the wrong of preferences that are adapted to oppression, however, is oriented around the question of how and when an agent’s preferences should be denied normative authority.[5] It asks when and why an agent’s apparent preferences might be ruled invalid, or of lesser normative weight, because of some higher-order or disqualifying consideration. Such questions overtly problematise the coincidence of the good with agents’ preferences, and it is thus imperative to consider whether an account of the good is required in order to ask them.
I separate attempts to explain the wrong of adaptive preferences into three broad approaches: appeals to internal conditions; to external options; and to external conditions. Observing their respective attitudes to an account of the good, and the way each responds to the critical failures of the previous, I chart a progression that ultimately reveals a distinct shift in the way that the relationship between preferences and the good is conceived. I will thus argue that differentiating the normative authority of adaptive processes as a political project is motivated by claims concerning the good, which themselves have normative implications.
1. Appeals to Internal Conditions
Approaches to the problem of adaptive preferences that identify internal processes or qualities as the site of the problem, do so with the intent of identifying the corresponding preferences as non-autonomous. This explanation relies upon the (uncontentious) premise that only on the basis of a preference being autonomously held is normative authority attributed to it. As such, non-autonomous preferences can be disregarded if this is what is required to promote the autonomy which the agent is otherwise lacking. This rests upon a second premise: that autonomy is not a wholly inherent quality, but one that can be both undermined and promoted. Insofar as internal approaches are all focussed around discussions of autonomy, then, they would appear to cohere with a value pluralism that justifies its respect for preferences through the nominal ‘good’ of autonomy.
I will address two paradigmatic definitions of autonomy adopted specifically for the adaptive preferences problem, with the aim of demonstrating their shared elements as the site of their shortcomings. Firstly, from Bruckner, who argues that adaptive preferences should not be differentiated from authoritative ‘autonomous’ preferences merely by virtue of their being adaptive. He then attempts to identify those that do not warrant the ‘presumption of normativity’ through a reflective endorsement criterion. [6] Since preferences are ordinarily, and beneficially, regulated by an agent’s conceivable and accessible options, lack of autonomy cannot be pinned to their contingency alone and must therefore be confirmed by a process of dis-identification after the fact. It is thus either (1) entirely the imperative of the agent to determine an adaptively formed preference objectionable, or (2) concluded from what would be preferred by the agent if the agent were to reflect. Although Bruckner claims a parallel for (2) with a condition of logical consistency (whereby an agent’s beliefs are consistent if examination would reveal them consistent), whether or not an agent would or would not endorse a preference is surely underdetermined until the agent actually opts to reflect. This, I believe, poses a bigger problem than he allows. An ‘external spectator’—who presumably rules an agent’s preferences worthy of endorsing based upon an external account of the good—is strictly prohibited by Bruckner’s approach. [7] Exalting the internal view of the agent, however, requires a default acceptance of all preferences until the agent reflects and dis-identifies. Given that preferences adapted to conditions of oppression are likely to be precisely those for which an agent has strong incentives to not dis-identify, this does not explain what is wrong with the preferences the feminist perspective needs to question.
Christman’s argument for the non-autonomy of adaptive preferences takes a different route to Bruckner’s by interrogating them on the basis of an authenticity condition. Christman argues that the problem can be pinned to the agent’s alienation from an operative practical identity. [8] The appeal of this account is the scenarios it excludes: when one agent alters her preferences to cope with the misfortune of circumstances such as a disabling accident, and another refuses to relinquish hers under oppressive conditions, the preferences can be understood to be autonomously held because in both cases they are formed out of the agent’s affirmation of her identity categories as reason-giving. [9] This pairing of cases outlines the stakes in allowing preferences formed in some circumstances to adapt unproblematically, while requiring others to be fully retained unadapted. Where Bruckner establishes a test for inappropriately adaptive preferences, Christman sets out what is normatively acceptable; both describe the conditions for preferences being truly the agent’s own, and thereby seek to avoid appeals to an account of the good. However, for Christman as for Bruckner, it is unclear that the authority by which the above two cases are determined autonomously-held preferences is not ultimately the ubiquitous ‘external spectator’—one who, in this approach, speaks for whether the appropriate identities are being affirmed.
Christman acknowledges that discerning the relevant aspects of an agent’s identity to adapting or retaining particular preferences faces a challenge of indeterminacy. [10] This concern, I believe, encapsulates two of Stoljar’s key criticisms of Christman’s approach. Stoljar firstly raises the worry, similar to the concern with Bruckner’s endorsement approach, that non-alienation could easily contribute to the motivation of an agent in accepting oppressive norms and thus functions as a shaky criterion for the normative authority of adaptive preferences. [11] The identity-affirmation approach must either appeal to a second order of acceptable motives for affirming various aspects of one’s identity, or else abdicate to the judgements of the ‘suspicious’ observer – as Christman ultimately does allow. [12] This, I would argue, is at the bottom of Stoljar’s final criticism that Christman may in fact need to import a normative framework to account for coercive influences after all. [13] It is telling that, in the concluding stages of his defence of adaptive preferences, Bruckner accommodates cases of oppression by doing just that. [14] Whether or not this secession undercuts his insights into adaptive preference formation, it certainly undermines his attempt to evade an account of the good.
Before moving to the next category of approaches, it is worth noting that one set of prominent internal-conditions positions has not been evaluated. These are theories of autonomy that posit a determinate quality—such as rational competence, or a sense of self-respect—whereby the agent’s reflective, identity-affirming, or other autonomy-implicating processes, are granted authority. [15] It will suffice to say for the moment that requiring internal processes to be characterised by a particular quality makes an appeal to the ‘good’ of that quality. I will evaluate the nature of this appeal alongside other direct appeals shortly.
2. Appeals to External Options
Appeals to external options present a possible rejoinder to the indeterminacy of internal approaches. They too seek to explain adaptive preferences in terms of autonomy deficiencies, but do so by construing autonomy as relative to the options available to the agent during preference formation. Elster, for instance, identifies the problem of an adaptive preference as lying in its being induced by necessity. [16] By defining proper preference formation in terms of freedom from necessity, adaptive preferences are disqualified in consequence of their not being true preferences. This is seemingly a neutral claim: a preference is considered problematic by virtue of its causally produced nature, thereby implying no conditions upon the content of those causes and no appeal to an account of the good. However, it faces the difficulty of an immediate incongruence with our experience of ordinary preferences, ruling invalid many preferences to which we might otherwise confer normative authority. Autonomous, normatively authoritative preferences would require complete self-sufficiency and freedom from all constraint, including biological. As Bruckner argues, even the keen tennis player who comes to prefer teaching, above competing as a result of her ability constraints, has her preference ruled non-authoritative on this account. [17]
Terlazzo presents an alternative to what she would call the failure in the ‘political efficacy’ of Elster’s account. [18] Autonomous preferences are those reflectively held in the presence of alternative options that are both recognised by the agent, and valuable to the agent. [19] This accommodates for the intuition that autonomy does not increase in proportionality to quantity of options; it is influenced to some extent by the value, presumably based on the agent’s subjective evaluation, of the available options. According to Terlazzo, the problem with adaptive preferences is that “they involve circumstances that either limit our awareness of the options that we have in some particular case, or make us recognise that the options that we would prefer are inaccessible.” [20] Adaptive preferences are thus attributed to a failure of circumstances rather than a failure of agential reflective capacities. Furthermore, this failure of circumstances appears prima facie to be nothing other than a failure in the promotion of value pluralism, and is therefore consistent with neutrality towards the good. However, Terlazzo’s recognition clause makes both this attribution, and this value pluralism, questionable. It is conceivable that an oppressed agent recognises no valuable alternatives to her preference, even though what might be perceived as valuable options are indeed available. Consider, for instance, Benson’s example of the young woman who spends inordinate time and effort anxiously sculpting her physical appearance, despite having ample other avenues available through which to pursue her own good. [21] What is it that makes this a failure of circumstances, and not the agent’s failure to recognise other options as valuable? Only if Terlazzo claims that, were the circumstances different, the agent would value other options, can the problem lie with the circumstances. For the same reason that Bruckner and Christman required it, the indeterminacy of this claim must appeal to an external spectator. Either Terlazzo allows blame to be focussed on the oppressed agent for a failure to recognise alternative values, or her explanation must appeal to an account of the good.
The care Terlazzo takes to avoid placing the burden of appropriate preference formation upon the agent draws out a key issue with approaches, briefly discussed earlier, that attribute these preferences to the lack of some internal quality. If such an appeal to an account of the good is truly only a marginal statement about an aspect of autonomy, then the problem with adaptive preferences is that the agent fails to have some right view of herself, or develop her own capacities appropriately. However, since Benson, Mackenzie, and others like them would surely not argue that culpability for the problem of adaptive preferences be the sole responsibility of the agent who holds them, their chosen internal conditions must be understood with reference to external conditions which at least contribute to their production. These theories thus imply the same appeal to an account of the good in the same way that Terlazzo’s does.
With Terlazzo, the approach to questioning the normative authority of adaptive preferences has undergone a distinct shift. Rather than staging an argument whereby adaptive preferences are not counted definitionally as preferences per say, she seeks to disqualify the content of a preference from being the authority on the agent’s good—and in such a way that the preference itself may feasibly still stand as a preference. Terlazzo states that “if a person’s preferences have been foisted on her by a hostile environment, then they may well not track her good.” [22] This invokes a different intuition from one that rules an agent’s good determined by their preferences, and that it is only not the agent’s good if it is not truly their preference. I hope to show in the next section that a politically effective concept of adaptive preferences is based in precisely this shift in assumption.
3. Appeals to Unjust Conditions
In contrast to an external approach that rests upon purportedly neutral notions of options, I distinguish a second external approach that makes a direct appeal to the nature of the external conditions under which preferences are formed. Khader and Levey argue that preferences formed in endorsement of oppression are not normatively authoritative because they are causally determined by oppressive conditions. [23] The problem, in this approach, is not the causally determined nature of a preference but the kinds of conditions to which its determination can be attributed.
Khader defines an adaptive preference as “a preference (higher or lower-order) that is incompatible with an agent’s basic welfare and is causally related to the conditions of oppression under which it is formed.” [24] For Khader, adaptive preferences are not defined in terms of autonomy deficiency. The justification of this detachment is of course contingent upon the definition of autonomy being applied, but the fact that Khader demonstrates it as a possibility is a telling indication that her concern with adaptive preferences is not whether they are truly the agent’s own. Indeed, she elsewhere argues that it is the nature of some of the more complex adaptive preference cases (such as those where an agent participates in a trade-off between bad options, or is a victim of ‘value distortion’) that they are “all instances of preferences from which we cannot easily read people’s ‘true’ or ‘deep’ desires.” [25] The ambiguity of an agent’s desires is not a hindrance to determining the agent’s preferences as adaptive under Khader’s approach. Instead, their adaptiveness is attributed directly to their being ‘deprivation perpetuating’. [26]
The nature of the causal connection between unjust conditions and adaptive preferences is left open by Khader. She lists a number of ways in which a preference might be the result of oppression, including that preference being held uncritically, under an inadequate sense of one’s self worth, or with limited awareness of alternatives. [27] The possibility that each of these characteristics might warrant denial of a preference’s normative authority in and of themselves, with no further appeal to an account of the good, has already been repudiated—and thus I argue that Khader’s approach makes explicit the appeal to an account of the good upon which they all rely.
The need for an explanation of adaptive preferences to appeal directly to the conditions under which they are formed is argued even more forcefully by Levey. She makes a compelling case that to define adaptive preferences as those formed under a lesser sense of one’s self-worth (or, I suggest, to some similar lack accorded to the agent) is to undermine the actual value of those preferences. [28] This is a separate problem to Terlazzo’s, and Khader’s, concern about disrespecting the agent. It is one that becomes particularly significant for contexts in which feminists might desire to attribute a distributary gender difference – such as the higher proportion of women who choose to work in the home in comparison to men – to adaptive preferences, without discounting the value of those roles. Levey acknowledges that often preferences that endorse oppressive conditions are formed in consistency with an agent’s ‘comprehensive conceptions of the good’, be that a religious perspective or cultural gendered value judgements. [29] To identify adaptive preferences through the unjust conditions in which they arise must ultimately ground their lack of normative authority in the conditions, and not in particular qualities inherent to the preferences.
Explaining adaptive preferences in terms of the conditions in which they are formed clearly relies upon an appeal to an account of the good. This account must include universal welfare conditions which, if unmet, are definitive of oppression. If the explanation also invokes a concept of autonomy (which Khader’s does not, but Levey’s does), then the good of autonomous agency is defined such that the agent’s beliefs about the good must be consistent with these welfare conditions in order to count as autonomous.
Khader’s and Levey’s approaches are, I would argue, the logical response to the shortcomings of Elster’s explanation in terms of necessity. They also follow inevitably from the shift in underlying assumption I have observed in Terlazzo’s approach. This shift can be explicated in terms of two contrasting claims upon which an explanation of adaptive preferences may be built:
- An agent’s good is determined by the agent’s autonomously-held preferences.
- An agent cannot autonomously prefer what is not conducive to their good.
Both (1) and (2) produce attempts to demonstrate adaptive preferences as non-autonomous, but while (1) makes no appeal to an account of the good, (2) presupposes an account for the claim to be verifiable. Bruckner’s, Christman’s, and Elster’s approaches are all committed to the value pluralism of (1). However, as has already been demonstrated, theirs and similar approaches fail to produce conditions for autonomy that match intuitions about problematically adapted preferences. I argue that this is due to intuitions regarding adaptive preferences ultimately being grounded in claim (2) and thus in the appeal to an account of the good that corresponds to it.
If (2) is true, then—presuming some relevant account of the good can be agreed upon—the normative authority of adaptive preferences can be straightforwardly denied. If the content of an agent’s preference is not to their own good, then it must not be an autonomously held preference. However, for reasons brought to light by Khader’s and Levey’s discussions of adaptive preferences, claim (2) cannot be so straightforwardly established. It requires, on the one hand, a non-neutral concept of autonomy which some, Khader included, have argued is best avoided—predominately for the sake of maintaining respect for persons. On the other hand, the assumption that an agent would never autonomously prefer something that does not cohere with his or her good itself requires justification. It requires Khader’s condition of causal relation to unjust conditions to be one of causal determination: it must always be the case that, were the conditions otherwise, so too would the agent’s preference. Levey’s acknowledgement of the prevailing concepts of the good, in which (adaptive) preferences are often grounded, is of cautionary significance here. Arguing that an agent is incapable of autonomously holding a preference that is not in conformity with the given account of the good, requires thorough justification in terms of a notion of human personhood, and presents worrying consequences for respect for agents’ preferences in general. If an agent’s preference is identified as adaptive through an appeal to an account of the good, then I would argue that this identification is a normative and not merely a descriptive claim. To deny the normative authority of such a preference is to claim that an agent should not prefer what is not conducive to their good. However minimal the account of the good upon which the identification of an adaptive preference is based, it must be recognised that this is a normative claim in just the same way that preferences held through appeal to gender norms—as might arise from an alternative account of the good—are normative claims. It may be a well-justified normative claim, but it is nonetheless not descriptive, and it is certainly not neutral.
Arguing from the politically ineffective indeterminacy that neutral accounts of adaptive preferences—whether based in internal conditions or external options—all collapse into, I have shown a progression of responses in Terlazzo’s, Khader’s, and Levey’s approaches that culminate in identifying adaptive preferences with the oppressive conditions they are formed under. All candidate qualities for the problematic nature of adaptive preferences refer back to these conditions. It is thus by appeal to an account of the good, conceived as an absence of oppression (which must itself be defined), that adaptive preferences are denied normative authority. Adaptive preferences, so theorised, are not discounted because they are not the agent’s ‘true’ preferences, but because they are considered to be preferences that an agent should not have.
Jemimah has recently completed her Honours in Philosophy at the University of Queensland with theses on Foucault and Descartes. She is fascinated by the philosophical frontiers of secularism and is currently undertaking some theological study before (hopefully) going on to do her PhD.
Endnotes
[1] John Christman, “Relational Autonomy, Liberal Individualism, and the Social Constitution of Selves,” Philosophical Studies 117, no. 1 (2004).
[2] See Donald W. Bruckner, “In Defense of Adaptive Preferences,” Philosophical Studies: An International Journal for Philosophy in the Analytic Tradition 142, no. 3 (2009).
[3] See Cass R. Sunstein, “Preferences and Politics,” Philosophy & Public Affairs 20, no. 1 (1991).
[4] See Natalie Stoljar, “Autonomy and Adaptive Preference Formation,” Autonomy, Oppression, and Gender (2014).
[5] Catriona Mackenzie, “Relational Autonomy, Normative Authority and Perfectionism,” Journal of Social Philosophy 39, no. 4 (2008): 512.
[6] Bruckner, “In Defense of Adaptive Preferences,” 312.
[7] Ibid., 322.
[8] John Christman, “Coping or Oppression,” in Autonomy, Oppression, and Gender, ed. Andrea Veltman and Mark Piper (New York: Oxford University Press, 2014).
[9]Ibid., 22-23.
[10] Ibid., 24.
[11] Natalie Stoljar, “Autonomy and Adaptive Preference Formation,” ibid.: 236.
[12] Christman, “Coping or Oppression,” 223.
[13] Stoljar, “Autonomy and Adaptive Preference Formation,” 238.
[14] Bruckner, “In Defense of Adaptive Preferences,” 322-23.
[15] See Paul Benson, “Autonomy and Oppressive Socialization,” Social Theory and Practice 17, no. 3 (1991). and Mackenzie, “Relational Autonomy, Normative Authority and Perfectionism.”
[16] Jon Elster, “Sour Grapes – Utilitarianism and the Genesis of Wants,” in Utilitarianism and Beyond, ed. Amartya Sen and Bernard Williams (Cambridge: Cambridge University Press, 1982).
[17] Bruckner, “In Defense of Adaptive Preferences,” 313.
[18] Rosa Terlazzo, “Conceptualization Adaptive Preferences Respectfully: An Indirectly Substantive Account,” Journal of Political Philosophy 24, no. 2 (2016): 210.
[19] Ibid., 215.
[20] Ibid., 220.
[21] Benson, “Autonomy and Oppressive Socialization,” 389-90.
[22] Terlazzo, “Conceptualization Adaptive Preferences Respectfully: An Indirectly Substantive Account,” 221.
[23] Serene J. Khader, “Must Theorising About Adaptive Preferences Deny Women’s Agency?,” Journal of Applied Philosophy 29, no. 4 (2012). Ann Levey, “Liberalism, Adaptive Preferences, and Gender Equality,” Hypatia: A Journal of Feminist Philosophy 20, no. 4 (2005).
[24] Khader, “Must Theorising About Adaptive Preferences Deny Women’s Agency?,” 210.
[25] “Identifying Adaptive Preferences in Practice: Lessons from Postcolonial Feminisms,” Journal of Global Ethics 9, no. 3 (2013): 320.
[26] Ibid.
[27] “Must Theorising About Adaptive Preferences Deny Women’s Agency?,” 310.
[28] Levey, “Liberalism, Adaptive Preferences, and Gender Equality,” 137.
[29] Ibid., 139.
Works Cited
Benson, Paul. “Autonomy and Oppressive Socialization.” Social Theory and Practice 17, no. 3 (1991): 385-408.
Bruckner, Donald W. “In Defense of Adaptive Preferences.” [In English]. Philosophical Studies: An International Journal for Philosophy in the Analytic Tradition 142, no. 3 (2017-03-21 2009): 307-24.
Christman, John. “Coping or Oppression.” Chap. 10 In Autonomy, Oppression, and Gender, edited by Andrea Veltman and Mark Piper, 201-26. New York: Oxford University Press, 2014.
———. “Relational Autonomy, Liberal Individualism, and the Social Constitution of Selves.” Philosophical Studies 117, no. 1 (January 01 2004): 143-64.
Elster, Jon. “Sour Grapes – Utilitarianism and the Genesis of Wants.” In Utilitarianism and Beyond, edited by Amartya Sen and Bernard Williams, 219-38. Cambridge: Cambridge University Press, 1982.
Khader, Serene J. “Identifying Adaptive Preferences in Practice: Lessons from Postcolonial Feminisms.” Journal of Global Ethics 9, no. 3 (2013/12/01 2013): 311-27.
———. “Must Theorising About Adaptive Preferences Deny Women’s Agency?”. Journal of Applied Philosophy 29, no. 4 (2012): 302-17.
Levey, Ann. “Liberalism, Adaptive Preferences, and Gender Equality.” [In English]. Hypatia: A Journal of Feminist Philosophy 20, no. 4 (2017-03-21 2005): 127-43.
Mackenzie, Catriona. “Relational Autonomy, Normative Authority and Perfectionism.” Journal of Social Philosophy 39, no. 4 (2008): 512-33.
Stoljar, Natalie. “Autonomy and Adaptive Preference Formation.” Autonomy, Oppression, and Gender (2014): 227-54.
Sunstein, Cass R. “Preferences and Politics.” Philosophy & Public Affairs 20, no. 1 (1991): 3-34.
Terlazzo, Rosa. “Conceptualization Adaptive Preferences Respectfully: An Indirectly Substantive Account.” [In English]. Journal of Political Philosophy 24, no. 2 (2017-03-21 2016): 206-26.
Featured image by Abby via flickr
EXORDIUM 