Autonomy and the Ownership of Our Own Destiny: Tracking the External World and Human Behavior, and the Paradox of Autonomy

Research on autonomy exhibits a constellation of variegated perspectives, from the problem of the crude deprivation of it to the study of the distinction between personal and moral autonomy, and from the problem of the role of a “self as narrator”, who classifies its own actions as autonomous or not, to the importance of the political side and, finally, to the need of defending and enhancing human autonomy. My precise concern in this article will be the examination of the role of the human cognitive processes that give rise to the most important ways of tracking the external world and human behavior in their relationship to some central aspects of human autonomy, also to the aim of clarifying the link between autonomy and the ownership of our own destinies. I will also focus on the preservation of human autonomy as an important component of human dignity, seeing it as strictly associated with knowledge and, even more significantly, with the constant production of new and pertinent knowledge of various kinds. I will also describe the important paradox of autonomy, which resorts to the fact that, on one side, cognitions (from science to morality, from common knowledge to philosophy, etc.) are necessary to be able to perform autonomous actions and decisions because we need believe in rules that justify and identify our choices, but, on the other side, these same rules can become (for example, as a result of contrasting with other internalized and approved moral rules or knowledge contents) oppressive norms that diminish autonomy and can thus, paradoxically, defeat agents’ autonomous capacity “to take ownership”.


Autonomy: A Constellation of Variegated Perspectives
To my knowledge, the research on the problem of autonomy is extremely rich and, at the same time, dispersed in a fantastic constellation of studies that, on only a few occasions, present the depressing features of scholastic proliferation of hyper-detailed analytic perspectives. Primarily, research ranges from cases in which the center of analysis is the crude deprivation of autonomy (cf. Räikkä and Varelius [1]) to the problem of moral and personal autonomy (cf. Giovagnoli [2]). autonomous and self-determined free choices. For example, in biosciences, the problem of how to enhance our given mental and physical capacities and of how to govern the biological processes that supervise normal aging have recently attracted the attention of researchers interested in the extension of human life. In the last moments of life, the new techniques of human enhancement are seen as ways of extending autonomy by freeing people from the constraints imposed by nature, even anticipating the future arrival of impressive physically and psychically immortal post-humans.
Furthermore, the problem of autonomy of disabled patients who reject life-sustaining therapy opens up various problems affecting the intertwining between autonomy and adaptation (cf. Räikkä and Varelius [1]): (1) Is the decision that guides people when they show control over the time and manner of their deaths autonomous? (2) Often, end-of-life choices occur in confused contexts in which their preferences seem to be affected in ways that jeopardize autonomous and voluntary choice.
(3) What about decisions concerning clinical trials for terminal illnesses, when, for example, a patient has to select between a harmless medication and a probably much more efficacious but hazardous experimental medication? (4) Are adaptive preferences connected with autonomy when wellbeing is promoted through the novel ways of enhancing human capacities that biosciences are expected to produce? (5) Many problems regarding autonomy are also opened by the adaptive preferences that are linked to situations of self-deception-facing death and serious diseases, people are inclined both to deceive themselves about their chances and, thus, to adapt their preference to what they consider conceivable for them. That is, are people completely autonomous with respect to the choices they adopt on the basis of adaptive preferences resulting from self-deception?

Moral and Personal Autonomy
Moral autonomy refers to the capacity to subject oneself to (objective) moral principles; on the contrary, so-called personal autonomy is morally neutral. Of course, moral autonomy strictly regards the relation between one person's realization of his own ends and others' realization of theirs, and so is ethically related to shared rules of behavior; when the word "morality" comes out, we are immediately dealing with a specific set of moral values of a given collective. Of course, from a philosophical perspective, a totally similar definition can be reached in the case of "legal" autonomy, in which the reference is still to the moral content mediated by the groups of laws of a given collective (for example, all of the citizens of a modern country).
Personal autonomy merely refers to those traits that single individuals can show relative to any aspects of their lives, not restricted to those related to moral duties; it refers to those people's activities not just regarding the following of their desires, but also in terms of the ability of choosing which of their desires to follow (cf. Christman and Anderson [3]). It is in the intertwining of this kind of autonomy with the moral one that we can conclude that a moral individual is autonomous in the moral sense when he is not only governed by his own ideas of happiness, but-to adopt a Kantian perspective-by a universalized concern for the ends of all rational persons (where the adjective universal refers to the moral frameworks shared by specific collectives of human beings, as I have indicated above).
Giovagnoli [2,4,5] usefully notes that a person is autonomous not only when he is capable of rationally recognizing reasons and motives for acting, but also when the responsibility of his choices is adopted for "subjective" reasons; autonomy requires the potentiality of "taking deontic attitudes" in order to make subjectively unequivocal and patent the material structure of the content of the reasons of our actions: In one sense, to be autonomous or self-determined is to be governed by the principles of your own causality, principles that are definitive of your will. In another deeper sense, to be autonomous or self-determined is to choose the principles that are definitive of your will. This is the kind of determination that Kant called "spontaneity". Every agent, even an animal agent, is autonomous and self-determined in the first sense, or it would make no sense to attribute its movement to it. Only responsible agents, human agents, are autonomous in the second and deeper sense. However, I maintain that an autonomous agent must refer also to substantive characteristics of the content of reasons for acting (cf. Giovagnoli [2] (p. 2)).
I contend that an agent can be called autonomous because he can take advantage, as I will explain in the following sections, of the results of human capacities to track the external natural and artificial world and other human beings' behaviors. We will see that the autonomous agent can be considered the fruit of both general knowledge and morality; how is it possible to be autonomous in our choices if we cannot base them on the predictive power provided by those knowledge and morality contents themselves? Only if we can count on correct predictions about events and human behaviors can we be endowed with autonomy.

Paradox of Autonomy
Which resorts to the fact that, on one side, cognitions (from science to morality, from common knowledge to philosophy, etc.) are necessary in order to be able to perform autonomous actions and decisions because we need to believe in rules that justify and identify our choices, but, on the other side, these same rules can become (for example, as a result of contrasting with other internalized and approved moral rules or knowledge contents) oppressive norms that diminish autonomy and can thus, paradoxically, defeat agents' autonomous capacity "to take ownership", as Giovagnoli correctly notes ( [2], p. 4). It is important to see that the autonomous choice (and its paradoxical character) emerges from an intertwining between social aspects (objective existent moral rules and knowledge contents), as a kind of objective space of reasons (cf. Brandom [6]), and individual aspects, which legitimate autonomous actions and decisions.
An interesting psychoanalytic perspective that also refers to another kind of "paradox of autonomy" has to be reported because it echoes-in a different intellectual framework (autonomy is thought as mainly referred to mental health)-the same emphasis that I devoted in this article to the fact that, to adequately grasp the concept of human personal autonomy, a reference to the intertwining of individual and social aspects has to be taken into account. Winnicott, in the article "From dependence towards independence in the development of the individual" [7], basically says that the concept of autonomy as "independence from others" would have to be-paradoxically-replaced by "independence with others": "In health, which is almost synonymous with maturity [. . . ] the adult is able to attend to his or her personal needs without being antisocial, and indeed, without a failure to take some responsibility for the maintenance or for the modification of society as it is found (p. 83)". In a sense, this paradox, illustrated by Cook [8], who repeatedly found it in her psychoanalytic practice, resorts to the fact that the achievement of psychological a healthy autonomy is dependent on recognition by another human as a kind of "constant tension between recognizing the other and asserting oneself" (cf. Benjamin ([9], p. 38)). Another important area of studies on autonomy regards the role of authenticity conditions and competency conditions in rendering actions as qualifiable as autonomous in the case of real agents. From a subjective perspective, the role of the so-called "self as narrator" (cf. Velleman [10]) is important. The self as narrator performs an abductive activity 1 devoted to formulating hypotheses about the status 1 Abductive cognition refers to all kinds of reasoning concerning hypotheses in both humans and animals. I have studied abduction in detail throughout my entire career, cf. my books [11][12][13]. of its own actions: human beings continually build and revise a cluster of self-accepted hypotheses that contribute to qualifying some of their own actions as autonomous or not. The self-deception (but also the time that passes) can seriously modify the self-perception of our own felt degree of autonomy of certain actions. We can easily consider cases where people's first-order drives and reasons better reflect their independent and sovereign personalities (their "proper selves", so to speak) with respect to the second-order ones, which may simply mirror persistent conditioning and spurious responses to social coercion in which autonomy is not seriously at work, even if felt as such.
The case of self-deception is even worse. Put simply, bad faith, as Jean Paul Sartre says [14], is a kind of falsehood that involves lying to oneself. In the human condition of bad faith, people treat themselves as means; they ignore or jettison the concept of choice in some respect because it is somehow vexing or burdensome, and in doing so, they relinquish autonomy and freedom and externalize responsibility. Once we fall into the condition of bad faith, we suffer a terrible consequence; bad faith denies us a full range of choice, erodes our ability to direct our own destinies, and, consequently, strips away a measure of our freedom and of our autonomy. Furthermore, this interplay between bad faith and diminished freedom, autonomy, and responsibility jeopardizes human dignity and leaves us more susceptible to exploitation and other forms of ill treatment. In my book [15], I contended that accruing our individual knowledge about ourselves and our condition can help us to manage our beliefs and guide us individually toward collective solutions to a number of difficult problems related to bad faith and its challenges to autonomy. 2

Autonomy and the Political Side
Another important category of studies concerning autonomy refers to the political side; for example, the obvious role of political liberalism in granting autonomy to citizens. Various topics are analyzed; for example, the claim that autonomy should not be considered as conciliable with certain constrained life circumstances-such as situations of social domination and self-abnegation-without reference to how-voluntarily-a person came to select or embrace that situation. Some researchers deeply discuss the so-called hyper-individualism of the liberal conception of the supposed autonomous person, together with the feminist critique of the typical masculine emphasis, which is embedded in this view; it would be conditioned by the excessive privilege attributed to separated and atomistic decisions. Indeed, they say, the emphasis on this atomistic type of autonomy does not have to conceal the role in it of socially implanted status of identity and value (as I already stressed in the previous subsections). 3 In the perspective from which I am presenting in this article, I will contend that, to grasp the concept of autonomy in a more complete way, a social and relational-intersubjective-self would have to be accounted for. On this issue, it is also interesting to quote the recently emphasized problem of joint and individual intentionality; we-intentionality (cf. Barber [16,17]) and personal autonomy in a social sense (for example, in habits and rituals) 4 indicate the problem of individual autonomy as a condition of a plural subject, and remind us that intentional autonomy does not preclude the so-called intentional hertararky, "a kind of other-regarding motivation that does not bottom out in one's own volitions (or motivations) and that does not reduce oneself to falling under the remote control of another or to acting as an intentional zombie" (cf. Barber ([16], p. 9)). Moreover, a study by Dragos [18], appropriately influenced by recent cognitive science research on distributed cognition and extended mind, introduces the so-called epistemic extension.
[. . . ] the idea that a subject can possess knowledge when other subjects possess some of the epistemic materials (e.g., evidence held, deliberations undertaken, inferences drawn, 2 On bad faith and self deception, see Chapter Five "Freedom and responsibility: Bad faith" of the book, quoted a few lines above, regarding the relationships between morality and our technological world [15]. 3 Cf. Giovagnoli [4,5]. 4 See the recent article by Giovagnoli [5]. cognitive abilities exercised) generating it. This second idea is opposed to the traditional tenet I call epistemic autonomy, according to which the possession of knowledge entails possession of all the epistemic materials generating it (p. 3).
At the level of social epistemology, group knowledge and group autonomy, when intertwined with this concept of epistemic extension, acquire a new perspective that takes advantage of the reference to distributed cognition and extended mind and departs from the conventional epistemological received view, merely refering to the concept of epistemic autonomy, also in the case of groups.

Enhancing Human Autonomy
When we pay attention to the problem of enhancing autonomy in human beings, we usually consider their embeddedness in a pluralistic society as the best ground, but of course, as "heirs" of the enlightenment, we also consider crucial the role of critical and self-critical rationality, capable of recognizing not only subjective reasons and motives for action, but also the objective ones available in a human collective. We consider a non-oppressive deontic social, moral, legal, and institutional framework as the best we can have for favoring human personal autonomy. Giovagnoli [5], to overcome this problem, which still resorts to the paradox of autonomy I have illustrated above, distinctly advocates an idea of "personal autonomy" in a social sense as the one that flourishes in discursive practices, but also at the prelinguistic or non-linguistic (model-based, cf. Magnani and Bertolotti [19]) level. She contends that when socialization generates oppressive rules, we can overcome them by promoting forms of living together capable of refusing oppressive norms at a political level, and favoring equality and trust.
In addition, taking advantage of the last ideas related to the problem of promoting personal autonomy, I think that a condition of possibility of this project appeals first of all to the examination of the role of knowledge in both tracking the external world (through common, philosophical, scientific, and many other kinds of knowledge) and tracking human behavior (through moral knowledge). These aspects are at the roots of the fundamental character of autonomy in its relationship with the ownership of our own destinies. I strongly think that a common but unexpressed thread of all the studies and perspectives illustrated in this and in the previous subsections can be found by focusing attention on the two roles of knowledge I have just indicated. All of the facets of the constellation of perspectives I have sketched in this section can receive further light from that point of view: In the following sections, I will indeed show that if we are interested in increasing human autonomy, the role of knowledge is fundamental. I will also contend that, especially in our technological era, knowledge is, consequently, a duty, also from the perspective of enhancing human personal autonomy.

Tracking the External World: Enhancing Autonomy Thanks to Predictive Knowledge
As I already anticipated in the last subsection, I think that everyday, philosophical, scientific, and many other kinds of knowledge about natural and artificial phenomena (as well as about the technologies that are linked to them) have permitted to humans a very extended collection of chances for choosing and acting in an autonomous way. I contend that the more one knows, the more relatively limpid alternatives one has, and one's decisions do not have to be affected by that uncertainty, which is generated by knowledge contents that make one's being autonomous very precarious. Indeed, the outcomes of a supposedly autonomous deliberator can instead be "decided"-so to speak-by other more or less unexpected variables, which qualify the autonomous character of one's decision as an illusion.
Thanks to the various kinds of knowledge that human beings are able to abductively generate, the natural and artificial external environment discloses two kinds of aspects: (1) The regular phenomena, which are such because they are made intelligible thanks to some kinds of knowledge results (for example, scientific), and are endowed with high degrees of predictability; this, in turn, renders possible a wide variety of free choices because we can count on precise expectations of some events and not of others; (2) unpredictable or unprecedented phenomena that are clearly seen as not within the reach of our intervention because we do not possess the knowledge that will permit their management-we consequently face more or less strongly partial autonomous decisions. Of course, there are no limitations in the possible construction of knowledge contents endowed with predictive power, also regarding these phenomena, because knowledge evolves and grows. In the presence of new knowledge contents of this type, a phenomenon that was previously treated as unmanageable, which is beyond the reach of human conditioning, can become, thanks to the new cognitive intelligibility, simply controllable, thus enhancing our choices options and, consequently, our autonomous free will. Our ancestors were not able to modify the course of a torrent, for example, because of the lack of the knowledge and the technical tools to do it, but new knowledge contents and new artificial instruments permit us to autonomously decide how to regulate them and to build artificial courses.
Following Dennett, we can call "elbow room" that "space" outside-out there-that forms the fundamental prerequisite for free will and autonomous choice (both at the individual and collective level). It is formed by (1) the fixed things, which do not present surprises and that are generally unnoticed, (2) the changing ones, which are trackable, that is, the ones that are known in a way that grants their predictability (at least under appropriate conditions), and (3) the chaotic ones, which are (de facto) unpredictable ( [20], p. 109), and on which we cannot reliably count. Dennett also usefully notes that chaotic systems constitute a "source of the 'practical' " because they often come as a surprise, and are thus an occasion of continuous chances and, of course, a source of potentially insecure decisions ( [20], p. 152).
It is the dynamics of human knowledge that make possible the individuation of those sorts of events and processes that Dennett calls "epistemic possibilities", which, in turn, are the condition of possibility for a human being to develop as an autonomous and free deliberator. These variegated kinds of events and processes depict what is "possible for all the deliberator knows or cares". They create the conditions for every "deliberator" (also an animal, of course) to be endowed with the appropriate information about their/its niches, conditions that that make them/it able to act in an autonomous way (or not) in them. We face a kind of epistemic openness, that is, the above "possibility for all one knows", which configures the "elbow room" I have introduced above. Of course, it is modern science that optimizes this epistemic openness by making available the great elbow room necessary for deliberation and, thus, the chance of efficiently tracking the external world, thanks to very extensive predictive abilities (also concerning the behavior of some technological entities-imagine, just as an example, drugs). When brain, consciousness, intentionality, and free will are working and a sufficient amount of knowledge (not only the one regarding the external non-human world; see the following section) is granted, we have many of the elements required to become markedly autonomous agents. However, we also need that kind of knowledge that refers to values and, thus, to moral contents, active in the related appropriate collective, 5 so that we can depict to ourselves our reliable goals and the potential good routes we might take because they will be in tune with other human beings' commitments. I have already indicated above that some of these knowledge contents are internal and some are external. 6 These external resources-new data about some potential choices, for example-can easily become internal when they are stored in memory and are, therefore, representable in our own brains. By picking 5 The reliability of moral knowledge is indeed circumscribed to more or less extended collectives, in which sharability of moral rules and behaviors is affected (and jeopardized) by both the presence of free riders and by the frequent changes of moral frameworks adopted by individuals, which create unexpected outcomes and thus a loss of autonomy. Something similar can be said in the case of general knowledge, whose reliability, contrarily to science, suffers from possible fuzziness and recurrent and more or less patent modifications. I treated the problem of the role and the dynamics of moral frameworks (and also their relationship with violence) in [21], especially in Chapter Five: "Multiple individual moralities may trigger violence. Engaging and disengaging morality". 6 The emphasis on the role of the external cognitive resources echoes the issues advanced by the view on autonomy in terms of the so-called epistemic extension, I have illustrated above in Section 1.5.
up new knowledge contents from external entities (a map, a moral sacred text believed by the members of our collective, or from the direct voice of another human being, for example) and by reproducing them in our brain cells, we can increase the variety of potential legitimate options and, so, we acquire more "room" to potentiate our free will and autonomy.
From a paleoanthropological perspective, the cognitive story I have just sketched can also illustrate the development of consciousness and intentionality from the ancestors to the current human beings (cf. Seth and Baars [22]): Consciousness and intentionality in human brains evolved in the way we know because human brains, in turn, have generated a lot of knowledge about the world, thus making available more and more external artifactual representations that, once re-represented in our brains, have hugely augmented consciousness and intentionality themselves. We know that consciousness and intentionality are the necessary condition for the development of free autonomous choice. In sum, we can hypothesize a kind of co-evolution between the evolution of consciousness-intentionality-free will-autonomy and of knowledge.

Creating Autonomy I: Tracking the External World through Scientific Knowledge
To become autonomous free choosers and actors that take advantage of the activity of tracking the external world, we certainly need everyday knowledge, but especially higher levels of knowledge, such as philosophy, science, and so on. An example of an external natural situation that-so to speak-does not enjoy rich predictive knowledge and that consequently makes voluntary and substantial autonomous choice impossible is the one regarding earthquakes. Seismology is a tradition of geological knowledge that, unfortunately, is not yet able to provide rigorous predictions: It is impossible to predict the exact moment of the next earthquake in a given area, even if seismologists can estimate some degrees of probability of the event. Consequently, we do not have any possibility to freely and autonomously choose when to go to a seismic area to visit certain relatives, being sure to be able to avoid the earthquake. Our capacity for free and autonomous choice is impeded; future new research in seismology will possibly be able to grant more rigorous predictions, even if it is well known that unpredictability is typical of natural chaotic processes, such as eruptions of volcanos or earthquakes.
I think I have sufficiently explained that, to become an autonomous and free chooser, an agent needs a rich amount of knowledge regarding the external world; in turn, knowledge, to be produced, needs conscious, free, intentional, and autonomous agents-an astonishing virtuous circle that characterized the evolution of human beings. In a dark, undetermined world, "out there", an autonomous and free agent is impossible. It is thanks to sophisticated human abductive capacities to generate good and reliable hypotheses that human beings have been capable of freeing the human organic bodies from the compelling pressures of their proximate environments. Currently, and fortunately, a lot of rich representations and inferential cognitive processes are present in the brains of beings-like-us, which can be exploited at will. Others are available out there, in the so-called cognitive niches, delegated to external mediators, where we can pick up the needed ones. 7 As I have already anticipated, to become free autonomous agents, however, we also need kinds of knowledge that are beyond the ones regarding phenomena of the non-human natural and artificial world. The external world is also pervaded by other human beings; if we are not able to predict their behaviors, it is impossible to also perform a minimally free and autonomous choice. Every choice could rapidly result just a fake choice, from the point of view of the status of its autonomy; you will rapidly find that you are not choosing, you are simply "chosen" by others' will and decisions because 7 Representational delegations are those cognitive human acts that transform the natural environment into a cognitive one. They are cognitive delegations to the external world that the mind itself has made throughout the history of culture by constructing the so-called cognitive niches. Humans have constructed voluminous cognitive niches, hugely endowed with informational, cognitive, and, more recently, computational processes, as illustrated by recent research in the field of sciences of evolution by Odling-Smee, Laland, and Feldman [23][24][25].
of your bad predictions about their thoughts, actions, and behaviors. You just made an "autonomous" decision in a Pickwickian sense.

Creating Autonomy II: Tracking Human Behavior: Rendering Human Behavior Predictable through Ethics
In the previous two sections, I maintained that cognitively tracking the external natural and artificial world gives rise to that "elbow room", which is fundamental in building a free and autonomous (individual or collective) deliberative agent, and that, regrettably, one of the main impediments to free autonomous choice is not only the absence of appropriate predictive knowledge about the functioning of the natural and artificial world, but also the absence of appropriate knowledge about the behavior of other humans. Moreover, other human beings are certainly "natural entities", but their thoughts, behaviors, and actions are very difficult to predict; for example: How can we track human intentions?
We know that the autonomy of our choices is linked to the aim of "morally" seeking own our own destinies; the autonomy grants that the choices are "ours" and that they are finalized to cause good outcomes through the exercising of our consciousness, free will, and intentionality. Of course, we can reach good results if we can count on some decency in predicting the behavior of other human agents, so as to avoid bad or dangerous surprises. Just to make an example, to try to "create" my destiny, when I was a young man, I always considered merit as a way of reaching important results, hypothesizing that my surrounding collectivity valued it similarly. I was wrong, so I partially failed to possess my destiny because my choices were free and autonomous, but only from a subjective, deceptive, perspective-that is, still in a Pickwickian sense. Actually, I have been considerably "decided" upon by external situations shaped by other human beings' behaviors that I was not able to correctly predict, and so I failed some goals that I thought I had adequately managed and envisaged. Many kinds of knowledge contents-for example, common and religious morality, moral philosophy, human and social sciences, and, of course, all other types of ethical knowledge-are patently related to our necessary exercise of conscious mental operations that shape our autonomous free will and the possible ownership of our own destinies.
How can I fruitfully be in an autonomous dimension if I cannot trust other human beings? How can I start an individual project or join a collective one if not by counting on an expected behavior of other human agents? We all perfectly know that religion, morality, moral knowledge, and education are finalized to build shared ways of behaving; this creates cooperation and, consequently, people's freedom and the possibility to adopt substantive autonomous choices. All the cognitive "tools" that are related to building cooperative collectives play the function of promoting substantive autonomy and free choice. To make an example, the practice of gossiping is not only an exchange of information about absent people with the aim of attacking them, but also a form of sociable interaction that distributes morally accepted schemes in a certain collective of humans by monitoring others as "moral characters" and building reliable hypotheses about what one can expect from them (cf. Yerkovich [26]). In the case of gossiping, the aim is also the one of (mildly?) policing free riders, that is, those people who enjoy the profits of social organizations but deny any kind reciprocation (cf. Dunbar [27]). 8 Kant said that the "Kingdom of Ends"-that is, the moral world-"is a practical Idea used to bring into existence what does not exist but can be made actual by our conduct-and indeed to bring it into existence in conformity with this Idea" ( [29], p. 104). Hence, the kingdom of ends is a kingdom of possible free and autonomous choices generated by (and contingent upon) human beings, for it is only their reliability that makes autonomy and free will, and thus responsibility and freedom, 8 Dunbar created the so-called "social brain hypothesis", which refers to the role played by the big brains of humans and primates in promoting not only information transfer, but also complex cooperative social systems; cf. my book [21,28]. possible. The following is the clear and explanatory comment given by Dennett about the so-called "self-made selves": Kant's famous claim in Foundations of the Metaphysics of Morals that the law we give ourselves does not bind us suggests that the selves we become in this process are not constrained by the law we promulgate because these selves are (partly) constituted by those very laws, partly created by a fiat that renders more articulate and definite something hitherto underdone or unformed ( [20], p. 90).
As I have illustrated in my book Morality on a Technological World. Knowledge as Duty [15], I think that the importance of knowledge is fundamental. I especially emphasized in that book that our era needs appropriate ethical frameworks that are able to defend humans' freedom, autonomy, responsibility, and the ownership of their destinies from the bad impact of technologies. We are indeed responsible for our own autonomy because its existence and its perpetuity depend on our intellectual and practical choices about knowledge, religions, scientific and moral institutions, and the related techniques and technologies, as well as on their use in everyday settings, work environments, education, communication, and economic life. That we can preserve, enhance, or jeopardize our possibilities of continuing to perform autonomous choices depends on our attitudes with respect to these factors. The example of a kind of paradox of liberalism related to the environmental problems can be of help to illustrate how complicated the intertwining between those factors and autonomy can be. Indeed, in matters of conservation of the natural environment, one could contend that neutrality is needed to respect the rights of the individuals, but this idea is evidently outbalanced by the fact that the freedom to jeopardize or dissipate natural goods and things today will, paradoxically, threaten freedom in the future, when our heirs will have, as a result, very few options when selecting and choosing among contending ideas of the good life.

Subjective Identification of Autonomy: An Abductive Hypothesis inside the Cognitive Self-Understanding of Our Own Actions?
In Section 1.4 of the first section, I illustrated the importance of the subjective "self as narrator" and its role in classifying its own degrees of autonomy. I said that these kinds of judgements are reached thanks to an abductive activity of formulating hypotheses about the status of our own actions. Indeed, human beings continually build and revise a cluster of self-accepted hypotheses that contribute to qualifying some of their own actions as autonomous and others not. I also said that, in this case, the self-deception (but also the time that passes) can seriously modify the self-perceptions of humans' own felt degrees of autonomy of certain actions.
More considerations on these aspects have to be added. Some neurological processes enable our brains to construct abductive explanatory hypotheses (also in the form of emotions) that "testify" that we are making autonomous actions. Gazzaniga [30] speaks explicitly of a "self-explanatory" activity-that is, abductive-of our minds when analyzing beliefs generated by the brain's left hemisphere interpreter. This portion of the brain monitors all the networks' behaviors and tries "to interpret their individual actions in order to create a unified idea of the self [. . . ]. It seeks explanations for internal and external events and expands on the actual facts we experience to make sense of, or interpret, the events of our life" (p. 148). 9 The self, by originating emotions and narratives about our degrees of autonomy (and of conscious and free will) is also part of the construction of a sense of identity, responsibility, and ownership of our own destinies. Starting at least from Italian humanism and Renaissance, human beings aim at seriously being in control of themselves rather than under the control of other human beings and/or 9 It is well known that in schizophrenia, however, the system of explanatory judgment that creates images of conscious will and autonomy is very different: Secret voices are "explained" as coming from the autonomous will of other selves. People are not able to track their own actual thoughts and, in many cases, also their own motor actions (cf. Wegner [31]). of natural and artificial entities: Faber est suae quisque fortunae. 10 As I have tried to demonstrate in this article, human beings themselves can augment, undermine, or even (through its considerable externalization) completely jeopardize personal autonomy. This occurs by administrating the processes of tracking the external world and other human beings. When humans favor and enhance that open "elbow room" they have disposed for themselves, which I illustrated in the previous parts of this article, autonomy increases.
In our present world, we face an increasing rate of interlacement of people with various external entities, especially of a technological nature. This hybridity tends to make it progressively more difficult to establish where the world stops and the person begins; of course, this fact constitutes a constant challenge to personal autonomy and to human freedom on the whole. 11 In our technological societies, obstacles and problems can flourish for a lot of reasons; for example, the absence of maintenance because of reduced budgets (typical of current neoliberal times), or the presence of incompetent workers on a task because of unethical hiring systems, and many other cases can be described (cf. [15]). In addition, computational systems have to be monitored to avoid their unpredictable taking possession of important technostructures that govern our lives, causing incorrect outcomes: Let us quote, for example, the electrical blackouts in huge areas in 2003, caused by a terrible mixture of computational inadequacies and work-environment failures. In this last case, millions of people lost various degrees of their autonomy to decide, and thus a considerable part of the ownership of their own destinies. It was a transitory reduction of that "elbow room" I repeatedly quoted above, created by humans exactly to grant their personal autonomy and the related free and autonomous choice. In addition, the loss of privacy is another important example of a threat to human autonomy. It is more than well known that computational technologies undermine it: In this case, autonomy is jeopardized because an individual is no more the only person that possesses important information about himself; the other human beings who know that information can easily exploit him to compromise the substantiality of his autonomous decisions. I cannot further expand upon the various problems regarding issues connected to the impact of computational programs on ethics and society or of other relevant technostructures. 12 In this article, I limit myself to dealing with cognitive and epistemological examination of the intertwining between the processes of tracking the external world and human behavior and the concept of autonomy.

Conclusions
In the first part of the article, I quickly described the main features of the current research on autonomy, from the problem of the crude expropriation of it, to the study of the differentiation between personal and moral autonomy, from the problem of the function of a "self as narrator" to define our own actions as autonomous, to the role of the political side, and, finally, to the need to protect and enhance human autonomy. With the help of the analysis of the human cognitive processes that give origin to various modalities of tracking the external world and human behavior, my main concern has been the study of the relationships between these processes and some central aspects of autonomy, also with the aim of shedding new light on the link between autonomy and the ownership of our own destinies. I also addressed the problem of the preservation of human autonomy as an important component of human dignity, as strictly associated with knowledge and, even more relevant, with the continual production of new (and appropriate) knowledge of various kinds. I also introduced the so-called paradox of autonomy, based on the fact that cognitions (from science to morality, from common knowledge 10 Every man is the artisan of his own fortune. Sallust attributes the quotation to Appius Claudius Caecus ("the blind"; ca. 340-280 BC), who was an important Roman statesman and dictator from a wealthy patrician family. 11 I have illustrated various problems concerning the relationship between technology and morality in our technological era in [15]. 12 Some more ethical details related to the current so-called "computational domestication of ignorant entities" are illustrated in recent articles of mine [32,33].
to philosophy, etc.) are needed to be able to perform autonomous actions and decisions because we need to believe in rules that justify and identify our choices, but, at the same time, these same rules can become (for example, as a result of contrasting with other internalized and approved moral rules or knowledge contents) oppressive norms that weaken autonomy and so, paradoxically, can defeat agents' autonomous capacity "to take ownership".