1. Gestation and genetics.

Think first of gestational surrogacy, where gametes from a man and a woman are united to form a fertilized ovum in vitro, which is then implanted in a second woman who is to gestate and give birth to the child, and then to deliver it to the genetic parents (or, perhaps, to other intended custodial parents, whoever they are). Assume that before the project is completed, the gestator threatens to withhold delivery unless she receives greater compensation. More profoundly, she claims the child as “her own” because she was the gestator—the birth mother—and that this outweighs the lack of a genetic link.

In such reproductive transactions, the gestational and genetic “components” of motherhood have been “split” or “unbundled”; the reproductive process, both physiologically and socially, has been fragmented and reassembled. There is no clear, dominating conceptual answer to the question, Who is the “true” or “natural” mother?—although some may believe that, given their particular evaluation of the physiological and social roles of gestation and genetics in identity formation, there is a clear answer.

Identifying the “natural mother” within the meaning of California's version of the Uniform Parentage Act was precisely the question in Johnson v. Calvert:³⁵

The problem reflected in the Johnson case is, in a plausible sense, a question about the “conceptual identity” of parenthood generally and motherhood in particular (is Mrs. Calvert or Ms. Johnson the real mother?), and about whether “motherhood” has to be “reconstructed” or “revised.” I do not have a rigorous account of what “conceptual reconstruction” means, but it seems unnecessary for present purposes. Perhaps “reconstruction” is in some cases too fancy a term: we may simply be “interpreting,” or clarifying what is implicit, or applying a given concept to new—but in this case very surprising—situations. In any case, when concepts are pressed beyond their apparent limits, their innards are exposed, revealing more of their strengths and weaknesses. This is an important payoff, however we describe our conceptual explorations.

The California Supreme Court did not resolve the matter by comparative genetic and physiological analysis of the workings of genes and gestation, or by inquiring into the possible exploitation of impoverished women of color (Anna Johnson was African American),³⁷

On that view, Mrs. Calvert, who was the genetic mother, was deemed to be the “natural mother” within the governing statutory meaning in that case, because that was how the transaction was set up: it was the Calverts' idea to create a family consisting of them and their genetic offspring, in close companionship.

If, on the other hand, the intention had been to assist the gestator in creating her own nuclear family, then, on the intention-is-determinative rationale, custody would have been awarded to her rather than to the genetic mother. This is not entirely unrealistic, especially if the gestator is a close relative or friend of the child's genetic sources.

The extent to which procreation is viewed as the intentional creation of a nuclear family as presently understood is a contingent social fact, though this view seems all but universal. To see this, compare human procreation as we know it with that in Plato's Republic,³⁸

This illustration of conceptual reconstruction or respecification emphasizes both the descriptive and evaluative features of this process: we are not just redefining abstractions, but making moral decisions on how to deal with novel situations. This is a very tricky business, since we are reviewing and revising some of the very tools needed for moral analysis in the first place.

2. Human reproductive cloning.

Human reproductive cloning produces a separate new person, as does traditional sexual reproduction. On that limited description, there is nothing new: the human population has increased by one. With sexual reproduction, however, we have conceptually (and morally and legally) clear ways to say who shall be the “default” custodial parents—that is, the parties automatically designated by law as custodial parents in the absence of evidence of a prior agreement on custody or judicial decree. The woman who brings a child to term is at least one of two default biological mothers. The sperm source is the biological father, period. But who would be the clearly required default biological parent(s) of the cloned offspring? Not the nuclear source, who is simply a previously born genetic twin. Not the woman who is the source of the enucleated ovum: she makes no genetic contribution (except for mitochondria, carried in the cytoplasm).³⁹

3. Mood control.

An example closer to our topic of trait change and identity reflects the increasing ability to control mental processes and to understand their neurophysiological substrates. Compare the four responses to the following question: “Why are you so energized and enthused—so inappropriately cheerful—all the time?” (a) “That's just the way I am; it's an aspect of my character; it's who I am as well as the way I am (which are, of course, connected).” (b) “I have naturally high serotonin levels.” (c) “I have naturally high serotonin levels, which largely accounts for that ‘up’ aspect of my personality—indeed, of my character—and helps makes me who I am.” (d) “I have naturally low serotonin levels, so I take a selective serotonin reuptake inhibitor, and now my serotonin is way up; I will remain on the drug indefinitely. It's the new me. Or, better, it's the real me that was previously suppressed by my inadequate serotonin chemistry.”

Are these descriptions and characterizations relevant to matters of personal identity? Are they relevant to assessment of character and actions? To the assignment of praise or blame? To freedom and responsibility generally? Does the example indicate the need for any conceptual reconstructions, whether concerning identity or any cognate concepts?

As for personal identity, no revisions are occasioned by modest degrees of coherent, continuous change,⁴²

However, the situation is less clear with the prospect of serious and rapid trait-malleability. Whether or not this impels revisions in our conceptual frameworks, it will change our expectations; our perceptions of someone's identity may become more tentative, driving us to distinguish between “real” or “former” identity and “induced” identity. Of course, the salience of new facts that erode our assumptions does not always require revision or abandonment of ideas (“identity and personal responsibility as we knew them no longer make sense”), but it may stimulate new conceptual and normative structures that adjust how we see and order things (“perhaps we should take more seriously the idea of diminished capacity”). Clearly, it is hard to distinguish between (on the one hand) the various overlapping ideas of conceptual modification, and (on the other hand) perceptual changes that involve revaluing or reordering aspects of what we see or adding “new” aspects. In any case, something is likely to change how we assess ourselves and our places in a world where nothing is nailed down, including our identities.

4. Identity itself: imaginable worlds of “shape-shifting”; what are the limits on the work “identity” can do?⁴³

I use the term “shape-shifting” loosely to cover all significant changes in whatever we view as an attribute, physical or mental/behavioral. Some uses of the term are thus analogical—as in altering the contours of mental functioning rather than specific physical shapes as such. The term “shape-shifting” is familiar in studies of mythology.

We do not generally think of discontinuous breaks in identity when maturity and aging take their usual courses; or when alertness is enhanced by caffeine; or even when steroids seem to generate visible results in appearance and performance. Still, even if we do not see breaks in identity, we might see “merit breaks” of sorts: Barry Bonds remains Barry Bonds, the baseball slugger, but those who believe he used steroids may downgrade the credit they assign to his otherwise impressive accomplishments (especially his setting the record for most home runs in a single season).⁴⁴

Nevertheless, it seems perfectly reasonable to posit far more radical changes—changes so salient within our customary⁴⁵

To see this, simply recall H. L. A. Hart's “philosophical fantasy,” as he put it, in which he asked readers to consider the fact that

Such “drastic” and “quicksilver” shifts in personal characteristics—which I take here to include changes in mental and behavioral dispositions as well as physical structure—would indeed seem to render “personal identity” irrelevant, unless perhaps we can identify some significant continuing invariant link across changes. But it is hard to believe we will soon, or ever, be faced with such a situation or its possibility—although I say this from our present perspective. And from that perspective, the role of “identity” in a Hartian world may seem compromised simply because it is no longer necessary or useful, or is much enfeebled.

Even in situations less chaotic than Hart's world, however, we might still be reduced to viewing identity (if we think of it at all), as a placeholder for, say, interpersonal and legal coordination—a sort of template or flywheel to maintain some stability through change. If Fx means x has the property F, we might view x as some continuing, ill-defined subject of change across all changes in properties (whatever F is); x is a continuing “repository” of (possibly) transient individuality—possibly a set of “successive selves”—although this still sounds too identity-oriented. We might also see F as a placeholder for the shifting properties that complicate our identity attributions. This two-placeholder conceptual concoction is, however, a rather thin template for our ideas of personal identity.

One might question the novelty of any of these Hartian or other imagined shape-shifting developments, and so question the need for conceptual retooling; the idea that there is nothing new under the sun serves to remind us that few developments have no “antecedents” whatever; at some level of abstract description, nothing is without precedent (except, of course, The First Thing). This is, however, a trivial point about the nature of abstract description. In any case, the illustrations suggest that some technological maneuvers will simply be too different to be comfortably subsumed within customary categories and practices.

1. In general; examples for continuing reference; some assumptions.

Explorers of the idea of personal identity sooner or later must ask about possible links and overlaps among the following:⁵¹

In an era of substantial technological governance over traits, however, I suggest one more consideration: when we move along any of these standard identity variables, we should take extra note of counterfactual identity—that is, supposed outcomes of transformative options not taken, including the option not to transform. Choosing among these options can occur at any life stage: the gametic, embryonic, fetal, and neonatal stages, infancy, and early childhood through adulthood. Such “alternative” identities might seem to some to constitute “real” identity, and may also affect perceptions of one's identity, by oneself and by others. Alternative identities may thus bear on what we view as an actual identity.

In reviewing some of these aspects and issues of identity, we can use as an occasional reference point the situation in Poul Anderson's science fiction novel Brain Wave:⁵²

A connected thought experiment (for future reference) involves somewhat different issues: deliberately induced memory loss targeted toward specific troubling memories,⁵³

It is now time to identify some important working assumptions in this essay: first, that a number of genes with “outsize” effects on intelligence and other major traits may be identified, and second, that it may be possible to work with them so as to significantly enhance the measures of those traits in the persons who will develop from an early embryo or from selected gametes.⁵⁴

2. Conceptual connections among identity, personhood, autonomy, and … ?

This investigation of inter-conceptual links was mentioned earlier. Ideas such as identity, personhood, and autonomy are conceptually joined in complex ways; perhaps they are each other in certain aspects. To say that one concept is an aspect of another concept, however, is not to say that there is really only one concept to identify; it is to link them logically. A distinctive feature of such mutually defining concepts is that they may refer to each other in an unending cycle. For example, whether a given trait change is thought to bear on identity may depend on the importance of the trait, but its importance may rest on ideas of personhood—say, on some attribute of cognitive understanding or perception. Are we dealing with separate concepts, extensionally equivalent concepts, or different ideas locked in conceptual ping pong?

1. The traits that seem to matter most to most people.⁶⁵

This discussion has parallels to that of “centrality” of traits, as discussed in Amélie Oksenberg Rorty and David Wong, “Aspects of Identity and Agency,” in Owen Flanagan and Amélie Oksenberg Rorty, eds., Identity, Character, and Morality (Cambridge, MA: MIT Press, 1990), 19–21.

The traits that matter most include merit attributes and/or resource attractors.⁶⁶

2. Characterizing the nature of the change: Coherence and continuity, now and in the future; intrusiveness; disorder and handicap; nature; impact on merit and blame ascriptions.

a. Coherence and continuity of trait changes; merit, desert, and distribution.

To some extent, “coherence” and “continuity” in individual lives and in personal interactions are needed if we want some degree of stability and predictability, and indeed they are required for many common social and political goals of human society. In turn, such stability depends on the relative constancy of the respective identities of individual persons. To close the loop, “incoherent” and “discontinuous” trait changes might be thought to compromise identity, and thus the stability of individual life paths and social interactions.

We invoke ideas of coherence in many contexts: judging scientific hypotheses, formulating theories of meaning, interpreting texts, evaluating developments in decisional law, and, more broadly, in assessing social relationships, societal goals, and personal life-plans. Like identity itself, coherence is an explicit or implicit constant or consideration across everything we think and do.

Analysts who rely on ideas of coherence often state that it is not a matter of simple consistency—as in requiring that a set of propositions not lead to both x and not-x. There is no reason to try to settle this. I observe only that it is at least arguable that whenever a conceptual system is said to be incoherent, that system—conceived of as a set of formulas or propositions—must at some point, however complex the path, lead to an inconsistency. However, demonstrating the inconsistency may be difficult or contested: a showing of contradiction presupposes interpretation, to put it loosely, and if there is no agreement on the meaning of the propositions whose consistency inter se is challenged, there is no closure on the issue of coherence. This is likely to be so with attributions of “incoherent” or even “discontinuous” trait changes: what seems to be a sharp change in behavior may be coherent, even smooth, in light of considerations that are not readily apparent. Coherence may be indeterminate because it rests in part on matters of perception and valuation, which in many cases must precede any search for formal inconsistency.

Ideas of (in)coherence may be important in distinguishing trait changes that are identity-relevant from those that are not. If any trait changes are indeed identity-relevant, they probably involve merit and resource-attractive attributes. In turn, this is relevant because, for better or worse, the ideas of merit and desert are broadly used to distribute the rewards of life generally. This holds to some extent in market systems: one's ability to compete for goods and services in such systems is often viewed as a (very) rough proxy for underlying merit.⁶⁷

b. Merit attributes and resource attractors in distributional systems.

Goods and services can be distributed through many different mechanisms. I note simply that outcomes of market transactions may be taken as rough proxies for merit of certain sorts, although the bare ability to command and bid resources for goods and services within a market can stand alone as a functioning (if not entirely sound) distributional criterion and does not require reference to “merit” at all.⁶⁸

Of course, markets and explicitly merit-based systems are not the only ways to distribute commodities. Some are distributed by, say, kinship relationships, conquest, theft, or competition within something akin to localized states of nature. However, at least in much of contemporary civilization, Western or Eastern, there is a rough correlation between characteristics viewed as morally praiseworthy merit attributes and those that can be described more neutrally as resource-attractors: intelligence and abilities of certain sorts; virtues such as reliability and loyalty; diligence and the capacity to exert effort. In this respect, markets are far more merit-related than, say, kinship systems.⁶⁹

These merit attributes and resource-attractors are historic targets for improvement and upgrading in traditional ways—e.g., “practice,” as the historic directions to Carnegie Hall go. They will certainly be prime targets for technological enhancement. Increases in the strength of one's resource-attractive traits may bring ever-increasing and ever-accelerating enlargements in one's economic, political, and personal powers. Increments in the very resource-attractors that brought in earlier resources continuingly expand one's “merit base” and thus one's bidding powers to command resources of all kinds in an accelerating positive feedback cycle. The distributional issues for merit-enhancing traits thus take on an importance lacking in other commodities that do not, by their very consumption, significantly raise the power of the distributee to make future merit claims for “merit-enhancing” (or any) resources—the very resource-attractors that allowed accumulation to begin in the first place.⁷⁰

Elsewhere, I refer to this as the problem of “Who merits merit?”⁷¹

c. The ways of change in attributes; personal betterment.

There may be limited exceptions, but most contemporary civilizations seem, in varying ways and degrees, to have absorbed the idea of progress, both for groups and for individuals: progress, in some sense, has become part of the canon of desirable, possibly morally mandatory human goals.

But personal progress has generally (if not exclusively) been understood to require effort, and perhaps luck. Advancement is expected to be incremental and gradual, and to require struggle, even for the natively endowed elite, at least when they are at the outer limits of their abilities.⁷⁴

Consider next the possibility of (partially) bypassing effort and using technological routes to higher attribute measures and skill levels. Popular commentary aside, there is an enormous difference between what present technologies of performance enhancement can do and what some people think they can do (or may do in the future). The public discussion of the use of various forms of steroids is an obvious example. Some people seem to think that ingesting them enhances strength or endurance with little or no effort, and the athlete becomes something that he or she “really” isn't. For that audience—especially the young and presumably more impressionable—the use of steroids may reinforce objectionable visions of free rides and bigger bangs for the buck. It seems far more accurate to say that ingesting steroids may have effects on physical (and possibly mentational) attributes at the margin, which may improve performance in small, if telling, increments. Small increments in performance may of course mark the difference between winning and losing, and require ever greater inputs of effort, with or without steroids. Although the time needed for any given degree of enhancement may be reduced compared with traditional efforts, one still doesn't just swell up with muscular bulk all at once.⁷⁵

Inflated beliefs or not, few think that anyone's identity is seriously compromised by using steroids. If Marsha wins Wimbledon and is found to have taken steroids, we don't say: “It wasn't Marsha who won, it was someone else, someone she turned into, like Billy Batson saying ‘Shazam’ and turning into Captain Marvel.” We say instead: “Marsha—our Marsha—seems to have won, but she really didn't win, not because someone else won (she-who-had-been-Marsha-but-isn't-anymore), but because she cheated.”⁷⁶

If we assume that steroid use had been “decriminalized” and proved harmless,⁷⁷

Next, given the technological and nontechnological (or less technological) routes to similar results,⁷⁹

d. Betterment and path dependence—the paradox of perfectionism;⁸⁰

I am using “perfectionism” loosely. The term has several varying (though linked) meanings. See generally Will Kymlicka, Liberalism, Community, and Culture (Oxford: Clarendon Press; New York: Oxford University Press, 1989), 33 (“A perfectionist theory includes a particular view, or range of views, about what dispositions and attributes define human perfection, and it views the development of these as our essential interest”). For a critical analysis of perfectionist theories, see Tom Regan, The Case for Animal Rights (Berkeley: University of California Press, 1983), 233–35. See also Ronald R. Garet, “Self-Transformability,” Southern California Law Review 65, no. 1 (1991): 121–203.

the relevance, if any, of what preceded bottom-line betterment.

This is how the “paradox” is structured: Consider the personal characteristics a society values the most at a given time—for example, intellectual abilities of certain sorts, physical strength and aptitude, and diligence in pursuing goals requiring strong efforts. These are the prime traits for improvement via traditional “upgrading” methods: study, training, and practice. Yet they seem to be the most untouchable with respect to technological alteration—a kind of tampering with the inviolable. Why inviolable? Perhaps because these traits are ordained by Nature or by God, or because the uncertain effects of rearranging basic life processes render it intrinsically (and at least instrumentally) immoral, or because of some vague precautionary principle that privileges nature and tradition as the presumptively safest routes. I have to leave this line of inquiry open, however.

The traits that we most want to develop, then, are also the traits that are to be the most insulated from artificial, unnatural, externally generated, and thus (possibly) identity-threatening augmentation. On this view, such augmentation would not be praiseworthy, would not result in enhanced merit, and might be affirmatively wrong. It may increase the resource-attractiveness of one's merit traits, but such an increase in resource-attractiveness does not, standing alone, really increase one's true moral merit; it only increases one's moral entitlements given compliance with the rules of the distribution system in place. (And perhaps whatever reward or credit is due must be shared with parents and bioengineers.) There is nothing intrinsically praiseworthy about merely enhancing such traits; the outcome is praiseworthy only if it is the terminus of the right paths. Improving ourselves in morally unsuitable ways thus does not perfect us; even worse, it seems to lessen us. In this sense, more is less. Or so one could argue.

The supposed paradox, then, is that (a) the merit traits most favored for enhancement are those for which technological means—the most efficient road to enhancement—are the most suspect; but (b) if they are indeed the most favored, they are probably the most resource-attractive and thus the most socially/economically useful; and (c) if so, we should want to enhance the measures of these most-valued traits in the most efficient way we can, so everyone can be better off through greater productivity, innovation, and economic advancement. There is thus a tension between maximizing (true) merit and maximizing socioeconomic utility, and in some cases, doing the latter defeats the former. The claim that technologized enhancement of merit attributes does not enhance or reflect merit constitutes one “arm” of the paradox, the other arm being our continuing interest in both merit and utility maximization.⁸¹

We can now say, tentatively, that technological augmentation at least suggests breaches of the integrity of personal identity—without altogether obliterating it—and leaves us confused when it is time to assign credit or blame and bestow awards or penalties. The process of technologized trait development may present results that seem incoherent or discontinuous with customary modes of self-improvement.⁸²

Moreover, to the extent that assignments of merit become dubious or even vacuous when we face technological enhancement, equality criteria in judging distributional schemes are also confused. On some theories of equality, the accumulation of greater rewards to the more meritorious not only does not violate equality standards, it is required by them. If we cannot tell who really won without cheating, how would we assign the reward? Who would be admitted to what educational institutions or awarded what positions? Should we just drop consideration of the person's prior trait profile and start dealing with the new one, however it was developed? Or, on the contrary, should we ignore the new configuration of traits and use the previous (now a counterfactual) one as the distributional guide? After all, we are, under current conditions, still dealing with the same person: despite the possible developmental incoherences and gaps between the before-and-after profiles, there remains a clear line of continuance that preserves overall identity.

e. Disorder control⁸³

For an exploration of the nature and limits of disorder and other models, see Norman Daniels, Just Health Care (Cambridge and New York: Cambridge University Press, 1985).

and personal transformation: Operating in the shadow of identity.

One of the moral claims embedded in a disorder model is that various nontrivial interventions⁸⁴

Many think it obvious that use of enhancement modalities “outside” a disorder model is wrong. It may be wrong, but this is not obvious. Consider the claim that “it wasn't really Jack doing the memory work—it was the medicine.” Medicine for disorder restores you to yourself by releasing your real self from captivity. Medicine for enhancement, however, takes you away from yourself and establishes a new, artificial plane of being and performance—or so it might be put. The moral monopoly of the disorder model is thus reinforced by these comparative accounts of causal attribution: in the one account, the self maintains its integrity while being restored by treatment—a process that is absorbed into an intact identity; in the other account, the self is transmuted into a new one by a process that remains “external” to the original self, however well integrated into the new.

Perhaps, then, there are realms of trait change in which there is a pathology that is identifiable, relatively discrete, and substantially removable without serious side-effects. In those cases, the removal of the pathology does not result in anything like global identity change. No one will say, “This isn't Sally”; they'll say, “This is Sally, flourishing, the new Sally who has the same identity as the old Sally but is better equipped for life because both her depression and her obsessive-compulsive disorder (OCD) have been controlled.” For such cases, it seems off the mark to apply psychiatrist Peter Kramer's phrase “better than well”⁸⁵

As for the role of identity: In Sally's case, perhaps the idea of identity, while instructive in many ways, does no significant obvious work. Still, our account seems to take place “in the shadow” of identity as an abstract descriptive and evaluative criterion:⁸⁶

3. Identity, person-perception, reduction, and objectification. Is cognitive psychology relevant? Trait alteration and identity change as risking reduction.

a. The idea of person-perception in cognitive science, and its connection to human reduction or devaluation; the bearing of objectification on identity.

In part, we perceive persons as persons—and as particular persons—by recognizing (perhaps not fully consciously) particular traits from the chaos of our sensorium. The nature of this process is contested both by scientists and by philosophers. But in whatever way it is accomplished, we see that this person is huge, that that one is rotund, and that the one over there is playing the piano. In many cases, we resist—for good reason—focusing on any given characteristic; we try to see the person as (very roughly speaking) an “entirety” (a gestalt perception of sorts), though with obvious emphases here and there. It is also clear that, at the same time, we often view persons as one thing or another, in varying respects. For example, here is Donovan the quarterback (or, sometimes, Donovan the black quarterback), and yonder is Martha the pianist. Most of the time, these are not viewed as reductive descriptions, and the people described would rarely protest them; the descriptions affirm their value as persons of strong talent and accomplishment.

Persons may not want to be described solely in these ways, however. They may fear that some forms of person-perception might be vehicles for reduction of a person's overall value to the value of a specific trait or group of traits (however “value” is understood in context). Donovan would surely insist that he has a life beyond quarterbacking, and Martha is not simply a pianist—and neither of them wishes to be viewed as a commodity.

Thus, we have another paradox (or irony), this time in person-perception: What seems to be one and the same perceptual process is intrinsically both part of an affirmative valuation of a person because of her favorable traits, say, Y and Z, and also part of her reduction to a merely useful entity because she bears those very traits, Y and Z.⁸⁷

“Reduction,” in this context, of course does not refer to claims that, say, all talk about mental entities is “reducible” to talk about physical entities and processes, or that, ontologically, one sort of thing (say, mental stuff) is really nothing but another sort of thing (material stuff). It is about human reduction—a psychological, perceptual, and evaluative matter. Full human reduction is the process or the product of viewing persons as having little or no intrinsic moral value; their worth, at least to persons other than themselves, is instrumental only. This instrumental value lies exclusively in a discrete trait or set of traits. I suppose the most obvious and total form of human reduction is slavery in its various forms, although one could argue whether the reduction was “complete” in any given setting.

In other forms of human reduction, however, persons are at least sometimes reduced to specific traits, in the eyes of particular persons or groups. To a customer, a prostitute might be “nothing but” her sexual attractiveness, sexual prowess, and her willingness to make them available commercially. To the owner of a professional sports team, perhaps the team members are viewed simply as mechanisms for assuring profit through the exploitation of their athletic traits.⁸⁸

The connection between this relational concept of human reduction and the often used term “objectification” is obvious.⁸⁹

If persons with certain traits—dwarfism, for example—become valued for purposes of, say, space travel in cramped quarters, then “breeding” them and impressing them into such service would be a paradigm of reduction of persons to nothing but (i.e., to being fully identified with) the value of a particular trait. Many refer to such reproductive “manufacturing” processes as “genetic diminishment,” but of course, relative to the tasks involved, the users of these persons—and perhaps the persons themselves—may view their traits as enhancements, not diminishments, at least when they are compared with others who wish to travel in space. It nevertheless seems clear that reduction and objectification are bound up with the notion of mere use, in the sense of the second formulation of Kant's categorical imperative, which enjoins us to treat persons as ends in themselves, and never as mere means.⁹¹

How do reduction, objectification, and mere use bear on personal identity, if at all? They would seem to affect identity as perceived by others and by oneself, and in some extreme situations the reduced/objectified/merely-used party may operate functionally as an object, from all viewpoints, including his own. To display the link to the idea of identity more specifically: Saying that “X has been reduced in value to her capacity to serve as a procreative vessel (e.g., a surrogate mother)” is saying, still more loosely, that she is “identified with” her reproductive capacities and her willingness to rent them out.

b. Planned identity change and the risk of reduction.

The very idea of directed trait change requires us to focus on the trait(s) in question: whatever drives this focus—self-improvement, enhancing the value of our slaves, ruling the world—is a matter of planning. If, for example, we have a reproductive plan at a fairly specific level of generality (compare “We want to have a child” with “We want to have a girl” with “We want to have a child with a genome identical to Meryl Streep's”), we are necessarily centering our attention on a subset of attributes (variously described) to the exclusion (at least temporarily) of others.

If technological intervention into personal attributes were serious or striking enough to raise identity questions, then the risk of reduction would probably be elevated, although not necessarily at a maximum. This is not part of an argument on my part against human enhancement; I think the risk of reduction/objectification/mere use is greatly overblown. It is not zero, however, and it seems likely that there will be instances in which directed trait change effects a reduction of one's perceived identity, and an elevation of one's perceived objecthood, both defined by the deliberately selected and altered traits.

I note one more perspective on trait change and its connection to reduction, objectification, and mere use of persons. Compare technological alteration of traits with the modifications of machines. If we are doing the same sort of thing to persons as we do to machines (compare adjusting the toaster from “toast” to “bagel” with switching Donovan the quarterback from “pass” to “run”), perhaps we are objectifying persons by assimilating—reducing—them to things.⁹²

When one makes comparisons, however, one is identifying differences as well as similarities. Upgrading a computer by installing more memory may bear an analogy to human memory improvement, but it bears fairly strong disanalogies also.⁹³

c. The cascading (perhaps chaotic) effects of directed trait change.⁹⁴

Aspects of this “impact analysis” are discussed in Rorty and Wong, “Aspects of Identity and Agency,” 19–21.

A trait change—particularly a sharp one—cannot be adequately analyzed without considering the relationships between that trait and others, and the complex network of effects deriving from the alteration. A significant change in appearance, for example (whether through plastic surgery, accident, disease, or aging), will have multiple, interacting, polycentric, and cycling effects that may be impossible to predict. And even the most effective and nonintrusive “magic bullet” treatments cannot just “subtract” or “delete” a disorder. When a patient's clinical depression departs, many things may change in complex ways. Moreover, with both enhancement and treatment, risks of error—however “error” is defined—are enormous. This is part of the reason why critics of enhancement technology tend to invoke “neophobic” preferences or “precautionary principles.”⁹⁵

d. A note on cloning and identity.⁹⁶

For a more extensive treatment of cloning issues, see generally Shapiro, “I Want a Girl.” I use the terms “clone” and “offspring” to refer to the person generated from the use of a somatic nucleus.

There are plenty of serious issues about whether we should engage in human cloning. Philosophical analysis of the idea of personal identity, however, is of no special importance within that discussion. The cloned offspring's traits are not altered from some previous form: the cloning process does not transform her any more than coming into existence transforms anyone. She is not an appendage of the nuclear source, does not share a consciousness with her, received no physical material from her beyond the molecules containing genomic instructions, and is not a “continuation” of her. She is genomically a delayed identical twin of the nuclear source. Nor has the cloned offspring appropriated or been forced to accept the source's identity as an individuated person. To deny this would entail an extreme form of genetic reductionism that few hold to: a person is not just her genome. There are, however, pragmatic issues about identity formation and person-perception in cloned offspring. Anyone who knows the story of James Mill's highly focused and rigorous training of his son, John Stuart Mill,⁹⁷

Still, cloning “magnifies” certain aspects of regular sexual reproduction, creating some puzzles for some senses of “identity.” Striking physical resemblance to the nuclear source is all but guaranteed. Genetically, a clone and her nuclear source are time-separated twins.⁹⁹

Given both the technological risks of producing gravely impaired persons and the risks of intrusive nurturing, some may believe that the prospect of harm to the clone renders the cloning enterprise immoral. But whether a clone is harmed turns on whether her life is, from her viewpoint,¹⁰¹

Although the risks to the clones are not best seen as affecting identity, the ills suffered may be serious nonetheless. Philosopher Joel Feinberg's notion of a child's right to an open future is a good talking point for assessing such harms.¹⁰²

If Mozart were cloned, however, one would expect the child to be at least introduced to music, and not left solely with the options of learning golf or studying philosophy. Indeed, trying to match a child's interests to her revealed or projected aptitudes is something that parents presumptively ought to do: such link-ups between interests and abilities promote the individual's opportunities, and thus, pro tanto, her autonomy, her chances for “success,” and her prospects for personal satisfaction in life.

1. In general; enhancement and its impact on awards, ownership, causal attribution, and identity; ascriptions of identity as moral claims.

If the concept of identity is a tool, what are the tasks, moral and legal, for which it is—or is thought to be—a tool? What descriptive and moral freight do our present notions of identity bear?

In a rough sense, the term “identity” seems to carry with it the gravitas of Nature, Fate, and God.¹⁰³

But just what are the underpinnings of this maxim? There is a link between what-is-natural-to-our-species (possibly including a self-improvement vector) and the idea of identity; it was not “unnatural” for us to come to the ideas of progress and perfection. Of course, determining why nature is a morally preferred standard of thought and action is yet another unbounded project. For our purposes, let us rest on this consideration: Nature may, in many cases, be the default or presumptive guide to what courses of action will “work” with reasonable efficiency, given our goals, and will produce a net benefit, all risks considered. This, however, is simply a loose “precautionary maxim,” not a deep moral principle.¹⁰⁴

Consider next the possibility of identity-challenging transfigurations. A working assumption bears restating: We presuppose continuity of personal identity through the inevitable changes of aging and experience, through the effects of traditional self-improvement, and through medical maneuvers accepted as treatments of disorders or repairs of defects or injuries.

But there are treatments and treatments: “I know Tommy had to have a new arm—but did they have to give him Koufax's?”¹⁰⁵

Earlier, I described the view that technological “improvement” might not enhance one's merit (although the project might be morally creditworthy for other reasons).¹⁰⁷

Finally, can ascriptions of identity be understood, at least in some cases, as moral claims? Think of these competing assertions: “The person (or creature) formed from the joining of brain structures from Marshall Artz and Tai Kwando is [more Marshall than Tai]; [more Tai than Marshall]; [both in equal parts]; [none of the above—we have a new person with certain memories and dispositions linked to the two earlier persons (have they ceased to exist?) and that's all we can say].”

Selecting the best of these claims is not a purely descriptive issue—but then, what sort of claim is it? What turns on it, and what does it turn on? If Artz was scheduled to inherit something from a recently deceased parent, a formal distribution must say whether the resulting creature is to take the inheritance because he is Artz, or is “the closest continuer” of Artz (or is this the same thing, exactly, as “being Artz”?). This is a question about the best description within a given social distributive system, and “best” clearly has morally evaluative components.

2. Hierarchies and gradients of traits.¹⁰⁹

Again, see the discussion of centrality of traits in Oksenberg and Wong, “Aspects of Identity and Agency,” 19–21. See also Nozick, Philosophical Explanations, 105 (“When people disagree about some of the problem cases of personal identity, for example, transfer of self from one body to another, this may stem from their different views of which dimensions, with what relative weights, constitute the metric of closeness. What is the correct measure of persons, then, and upon what does its being the correct one depend?”). See also Heyd, Genethics, 163–64.

When transformations—technological or otherwise—occur, how are they to be dealt with by the persons transformed and by others, within the framework of the reigning normative system? How are they characteristically dealt with now? (We do, from time to time, catch people having epiphanies.) The answers to these questions rest partly on the nature of the transformed trait—its “centrality” or importance for given purposes—but in whose view and on what standards? The views of the transformee, of the persons she interacts with, and of others who observe or know of her may all differ inter se.

Much depends on the relative importance of the trait for normal species functioning and for various culture-relative tasks and expectations; the nature, degree, and scope of trait change; the suddenness of change; its continuity or coherence with the trait as it was, and with coexisting traits; its cascading, and perhaps chaotic, personal and interpersonal effects; and so on.¹¹⁰

Finally, I offer one more illustrative question about how to sort and rank traits and trait changes by assessing their impacts on personal identity: How are we to view “gender transformation” procedures? Defining what this means is difficult: the idea of genomic changes in existing persons seems far-fetched, and not everyone views the results of the surgical/hormonal procedures involved as gender changes. Still, for those who view gender largely as a matter of social/legal “construction” or perception, the continuation of certain unchanging genetic or other physical variables does not block acceptance of gender transformation. In any case, gender transformations as now understood obviously involve trait changes of various sorts, and the personal identity issues that arise are obvious: gender seems universally of central importance to identity as resting on self-perception and perception by others. For those whose interior vision compels them to view themselves as “a woman in a man's body” or the reverse, gender transformation might be viewed as an “enhancement”—or perhaps as “relief of a handicap,” or even “amelioration of a gender-dysmorphic disorder.” Whether the physical interventions should be seen as generating a “biological” sex change or a “perceived-sex” change or any other sort of change is a hard conceptual problem that I do not address. Whatever form the intervention takes in particular cases, however, it may not be clear whether we should say, or deny, that a person is still whoever he was even after his gender has in some sense been transformed.¹¹¹

3. “Identity models” and “disorder models”: Can they properly authorize and limit technological alteration of traits? Can they operate concurrently?

At a high level of generality, the abstract logic of an identity-preservation model is identical to that of disorder and enhancement models: to establish authorizations and limitations on what we may, must, shouldn't, and mustn't do in modifying our traits.

Here is a very simplified account of what might be intended as a disorder model,¹¹²

A disorder model is thus a set of rules, principles, and standards for sorting the conditions we observe (and infer) into disorders, diseases, injuries, defects, etc.; for making specific diagnoses; and for authorizing and limiting certain forms of action (treatment, dealing with persons in certain ways when they are “ill”—i.e., installing the “sick role”—and so on). The key conceptual ingredients of such a model are highly contested in various situations. There is, for example, the view that specific disorders are often “constructed” in response to available technological fixes—or, more pointedly, that disorders are simply invented so that certain medical/surgical procedures can be used.¹¹⁴

What is “intrusive” is not a matter of pure empirical and conceptual analysis: it is heavily influenced by custom and other situational variables, and it is in many cases, at bottom, a moral characterization. Compulsory education throughout the years of one's childhood and adolescence has massive effects on one's development, but is not generally compared in “intrusiveness” to brain surgery used to reduce maladaptive behavior.

Handicap models are structurally isomorphic to disorder models but have quite different contents. For example, in the absence of pituitary or other disorders that impede normal growth, some physicians will not prescribe human growth hormone: its use is not authorized—and is implicitly banned—within a disorder model because it is pure augmentation.¹¹⁶

As an action-evaluation tool, an identity model implements the claim that “doing X (to) (for) (on) person W is justified for enhancement (or other) purposes only if it does not risk changing his identity (or his identity with respect to trait μ).”¹¹⁸

All of these justificatory models require their own justifications. Attempting to provide them often involves use of a family of morally freighted categories for determining the proper occasions and limits for modifying ourselves, temporarily or permanently. There may be overlapping appeals to what is natural; to what is internally rather than externally generated; and to what is authentically “of or from” the person—an aspect of causal attribution distinguishing “self”-agency from outside determinants. Some of these concepts are taken as bases for pejorative (if hyperbolic) ascriptions (“It wasn't you, it was the steroids that won”) or for excusing conditions (“It wasn't you, it was the [booze] [depression] talking”—or even “It wasn't you, it was your medicine that committed the murder”).¹²⁰

Legal issues of “synthetic sanity,” mentioned earlier (at the end of Section I), roughly illustrate some of these problems in moral ascriptions relevant to questions of personal identity. As we saw, it is possible that an insanity defense in a criminal trial would be weakened if the accused is medicated in order to render her competent to be tried: if her symptoms indeed remit with treatment, she will not seem to be crazy. Instructing juries to disregard her “induced sanity” at trial in judging her insanity when the act was committed—and thus her culpability—is unlikely to be an adequate corrective.¹²²

In many ways, then, these jurisprudential problems are identity (or identity-connected) arguments: to recognize a person as having a particular identity is to establish a template for determining what she, as opposed to someone or something else, has done or avoided. The insanity defense itself, for example, is implicitly bound up with notions of identity as one's “true self”—a true self that is overwhelmed by an “external,” “non-self-related,” identity-alien sickness. Identity, moral agency, and causal attribution are thus logically linked, at least in this respect.¹²⁴

There thus seem to be important intersections among identity, disorder, and enhancement models, beyond their obvious links as rough gauges for judging interventions into body and mind. In many situations, the models reinforce each other; indeed, perhaps they are connected elements of a larger model, or are aspects of each other. For example, some interventions not justified by disorder because they “go beyond” the applicable therapeutic logic (Tommy John's new pitching arm?) may also fail within an identity-preservation model because the induced changes are too discontinuous and radical.¹²⁵

Identity and disorder models are generally (though not inevitably) arrayed against enhancement models, which purport to justify serious intrusions at least when they are exercises of the subjects' or their parents' (or proxies') autonomous choices, or because of the expected gains from modification. Suppose, for example, a substance that provides one with eidetic perceptual powers for an extended period carries serious risks—say, subsequent memory impairments, blood-pressure spikes, and purulent acne. The reasons for declining this trait-change might not rest exclusively, if at all, on memory enhancement as an outcome; the main reason might be to avoid the risks in accord with a maxim telling us not to ingest dangerous substances unless there is a disorder-based reason for it. In contrast, an enhancement model might specify that a significant improvement in an important trait might be a strong reason to run the posited risks. And embedded in these considerations is the possibility that the acquisition of the powers of eidetic imagery might have transformative effects on one's capacities, behavior, and self-image. For some, this may either weigh against the intervention (an identity-preservation model) or pull for it (an identity-seeking model).

Identity, disorder, and enhancement models may thus enjoy an uneasy coexistence in which they may reinforce or limit (or have no effect on) each other. Each model helps call attention to the strengths and vulnerabilities of the others. Disorder and identity models may form pragmatic standards for limiting what might be unchecked distributional pressures generated by enhancement models—or autonomy models ungoverned by notions of disorder, identity, or enhancement. They may also be simultaneously invoked or challenged when particular interventions are in question (e.g., “gender transformation”). Here, issues of treatment of disorder, remediation of handicap, gender identity and its bearing on personal identity, and enhancement (should gender transformation be withheld from persons not suffering from a “transsexual disorder”?) are all immediately present. Of course, all of the models are embedded in complex regions of general moral and political philosophy from which we can raise still deeper issues: on what moral theory should disorder or identity-preservation ever limit autonomous self-change?

Finally, for some observers, disorder and identity models, insofar as they limit enhancement models, are valuable primarily to vindicate principles derived from the moral force of nature. Although nature's moral standing does not rest exclusively on religious premises, it holds a rather higher station within many theological systems than it does in secular thought. To be sure, some views of nature, secular or religious, may support enhancement models, but this does not seem to reflect currently dominant views.

4. What is the relevance of alternative identities not chosen? In particular, what is the relevance of the fact that one's parents rejected one's default natural identity, or any alternative identities?

What is the significance of counterfactual or, more particularly, default natural identity? First, knowing of it may affect a person's self-image, both with respect to identity and with respect to his sense of autonomy, individuality, and uniqueness. Second, knowing of it may affect the views of others who know of the now nonexistent but once possible identity.¹²⁶

To illustrate some of these risks, consider this remark from the President's Council on Bioethics, referring “to the question of what it might mean for a child to live with a chosen genotype: he may feel grateful to his parents for having gone to such trouble to spare him the burden of various genetic defects; but he might also have to deal with the sense that he is not just a gift born of his parents' love but also, in some degree, a product of their will.”¹²⁷

On the surface, at least, the problem described does not seem to raise issues of personal identity. It suggests risks of overly aggressive parenting in fulfillment of the plans of the reproductive enterprise, and risks to the child's sense of autonomous individuality—his belief that he is someone whose preferences are authentically his, rather than his parents'. It may also inspire fears that his parents' bonds to him were and are contingent on their reproductive/enhancement plan's favorable outcome.

But shifting the genome of an early embryo or a gamete from what its basic developmental fate might otherwise be may indeed bear on perceived identity, at least from the viewpoint of the parents and other external observers. What would be the point of intervening if not to generate benefits from it—for the child, for the parents, and possibly beyond? And how would one promote these goals without careful implementation that may heavily affect the child's development and ultimate “identity”? Planning drives realization.

It remains, however, that such programs are unlikely to generate lives not worth living from the viewpoint of the offspring, despite the emotional impact of learning of one's not-fully-natural origins and of one's abandoned default identity, and despite the risks of excessive constraints being imposed on the offspring's lifestyle and life path.