General rules and the normative dimension of belief in Hume ’ s epistemology

The main concern of this paper is whether Hume’s account of belief has a normative dimension, especially concerning his account of general rules of reasoning in his Treatise of Human Nature, and consequently, whether it is possible to offer an account of the normative force of those rules in spite of his naturalist framework. I conclude that there are many normative elements in his conception of belief and reasoning, and that, as many authors in recent studies of normativity have suggested, naturalism can sufficiently account for the normative structures of our cognition and their normative authority. Such a view of the normative dimension of belief in Hume’s epistemology also shows an interesting and close connection with the moral dimension of his thought, which I believe is of fundamental importance for understanding his thought in general.


General rules and the normative dimension of belief in Hume's epistemology
As regras gerais e a dimensão normativa da crença na epistemologia de Hume In recent decades, Hume's ap arent concern for normative issues in his philosophical writings has been of increasing interest among scholars (see Hearn, 1970Hearn, , 1976;;Martin, 1993, Fa kenstein, 1997;Lyons, 2001;Searjeantson, 2005;Mor is, 2006;Schliesser, 2007).One of the main questions seems to be whether his account of general rules in his Treatise and subsequent writings is an expression of normative claims concerning epistemic and moral judgment, or, on the contrary, whether ap ealing to rules is nothing more than a careless use of normative lan uage within a naturalist framework, which cannot provide a foundation for the normative force of epistemic rules (see Lyons, 2001, p. 270, 273, n. 14;Fa kenstein, 1997, p. 30).
Even from the very beginning of the discussion one could ca l the matter a "pseudo pro lem" based on an anachronism.In fact, the concern with rules was common throughout early modern philosophy, at least since Descartes's Re ulae a directione ingenii, and e ecia ly in early modern logic2 (see Easton, 1997;Serjeantson, 2005, p. 188), while normativity is itself a 20 th -century concern of philosophers.
In this paper, I wi l be dealing with some difficulties concerning how to articulate a normative view of Hume's account of general rules.In order to do this, I wi l try to support the fo lowing three claims: (i) Hume uses the concept of rules in at least three different ways, one of which is normative in a strong sense.To sup ort this statement it wi l be necessary to reca l what a general rule is and to examine its relation to belief and normativity; (ii) If there is a source of the normative force of general rules in Hume's thought, then this means that the theory is in itself normative, even though neither the general rules nor their normativity are explicit subjects of analysis in the Treatise, but instead serve as operative concepts; and (iii) Hume's novel view of philosophy and reason explain to some extent the generation of normative structures in his philosophy.

Rules in Hume's philosophy
That Hume's thought is concerned with the problem of rules in its central parts has been we l known since Hearn's two papers on general rules from the 1970's, in which he shows that general rules play a systematic role in the Treatise, being present in each of its three books (Hearn, 1970, p. 404-406 3 ).In this paper, however, I wi l be dealing only with the pro lem of the normativity of general rules in Hume's epistemology, mostly in part 3 of book 1 of the Treatise.There we find for the first time an extensive exposition of general rules and their influence on our judgment and belief.
General rules are, as described by Hume in T 1.34 , generalizations concerning the behavior of a kind of phenomenon of our experience that can be expressed by proposition in the form "every X is/has the property/ is predica le-of Y" .This kind of general statement ap ears within the analysis of probabilities, for they lack the universality of mathematical and logical propositions and they cannot be demonstrated (as Hume understands the term, probabilities are both our beliefs of matters of fact, as we l as the statements themselves that articulate such beliefs).Instead, they are conveyed by the imagination's tendency to generalize, based on past experience and custom.Nonetheless, not every generalization has the same status, and this fact reinforces the distinction commonly drawn in the secondary literature between general rules of prejudices (also refer ed to as "extensive general rules") and the so-ca led cor ective general rules.
The way in which general rules affect our judgment is also a dressed by Hume in his treatment of probability in the Treatise since it also belongs to the topic of belief and belief-formation mechanisms.According to Hume, a belief is a "strong and steady conception of an idea" that includes a claim to truth and with a number of different causes, such as memory, imagination, and causal inference: "we join belief to the conception, and are perswaded of the truth of what we conceive" (T 1.3.7;SBN 96-7 footnote).Hume natura ly endorses some of those mechanisms which are in a better position to fulfi l that expectation (see Loeb, 2002, p. 13) and prevent our ideas from being the mere "offspring of the imagination" (see T 1.3.9.4; SBN 108).Besides, considering the probability of causes, Hume holds that our judgments take place by virtue of custom and general rules (see T 1.3.12.24;SBN 141), and that "custom can lead us into false comparison of ideas" (T 1.3.9.17; SBN 116.See also T 1.3.13.2;SBN 143-144), e ecia ly when we, as a result of the imagination' s propensity to generalize, form general rules of the fo lowing type: "An Irishman cannot have wit, and a Frenchman cannot have solidity" .This kind of judgment is ca led an "unphilosophical ecies of probability" and "is that deriv' d from general rules, which we rashly form to ourselves, and which are the source of what we properly ca l prejudice" (T 1.3.13.7;SBN 146; for another example see T 2.2.5.12-13;SBN 362).This first kind of general rule leads to false reasoning in so far as the rule is caused by the "propensity of the imagination to extend the scope of judgments formed in one set of circumstances to other resem ling but non-identical circumstances" (Hearn, 1970, p. 405).
Although every judgment on probability is a function of custom, i.e. past experience and the projection of the imagination, the mind is not condemned to fo low its faulty natural tendencies.It is possi le to prevent the mind from forming false beliefs founded on rules of prejudices and from "the reposing any assurance in those momentary glimpses of light, which arise in the imagination from a feign'd resemlance and conti uity" (T 1.3.9.6; SBN 110), by means of the reflective mediation of second level judgments.5Thus, the same propensity of the imagination to generalize can result in "philosophical probability" when it is mediated by reflection.Reflection is a pivotal element in Hume's account of mental a ivity, it explains some a ects of our cognition by playing a twofold ro l: transforming some instances of the generalization tendency of the imagination into patterns of adequate judgments, i.e. proper "general rules"; but it also distin uishes, by means of cor ection, between those cases that are in accordance with those rules.That is why Hume su gests that mediation in judgment leads to the so-ca led cor ective general rules, which are a lowed "to influence their judgments <of men> even contrary to present observation and experience" (T 1.3.13.8;SBN 147, clarification a ded).In a very central passage for this investigation Hume claims that We shall afterwards take notice of some general rules, by which we ought to regulate our judgment concerning causes and effects; and these rules are form'd on the nature of our understanding, and on our experience of its operations in the judgments we form concerning objects.By them we learn to distinguish the accidental circumstances from the efficacious causes [...] The general rule is attributed to our judgment; as being more extensive and constant 6 (T  1.3.13.11;SBN 149, emphasis added).
Hume continues: "Sometimes the one, sometimes the other prevails, according to the disposition and character of the person.The vulgar are commonly uided by the first, and wise men by the second <kind of rules>" (T 1.3.13.12;SBN 150).It seems that the character of the beliefs a person forms, that is, the way someone structures his or her doxastic life, reflects the extent to which he or she is actua ly influenced by epistemic norms, the extent to which his or her beliefs express rationality or ir ationality. 7Thus for Hume, the rational epistemic agent is the one who is a le to assume a critical philosophical perspective.A wise person is someone whose beliefs are relia ly formed due to a relia le disposition to judge reflectively, and justified for the same reason (setting aside the pro lem of the criteria for justification of belief ).This is because, according to the cor ective general rules account, a rational belief not only expresses a healthy mental attitude of a believer, but it is also somehow related to the content of the beliefs.8Thus, according to Hume, more extensive and constant experience is "of a grosser and more stu born nature, less subject to accidents, and less influenced by whim and private fancy" (Essays9 1. XIV, p. 112; G&G, p. 175). 10An extended and constant experience of the same phenomenon or kind of phenomenon in the light of certain evidence (copy principle) is of a nature that can be expressed by general rules: "But however intricate they may seem, it is certain that general principles, if just and sound, must always prevail in the general course of things, though they may fail in particular cases; and it is the chief business of philosophers to regard the general course of things" (Essays 2. I, p. 254; G&G, p. 287) 11 .
As we can see above, there is an underlying connection between what can properly be an object of our knowledge and the method that can leads a reasone to that knowledge; a connection that we can make sense of only by ap ealing to Hume's account of general rules.Along this path, I ar ue that in order to fu ly understand the aforementioned connection, it is necessary to distin uish not two (as has been often done), but three different categories within the concept of "general rule" .Firstly, there are extensive general rules of prejudice.Secon ly, there are general p inciples, which are, so to eak, materia ly determined, for they express ecific properties or chara eristics of phenomena (in physics, politics, economics, moral, for example) and cor espond to the distinction between "what is owing to c ance, and what proceeds from causes" (Essays 1. XIV,p. 111;G&G,p. 174;also T 1.3.11.12; or, according to Hume, "between particular deliberation and general reasoning" (Essays 2. I p. 254).The conditions for the achievement of this kind of knowledge12 -which cor esponds to Hume's philosophical probability -are "the greater refinements and improvements of human reason" (Essays 1. XIV,p. 118;G&G,p. 180).This also sup orts Hume's claim that politics, metaphysics and morals "form the most considera le branches of science.Mathematics and natural philosophy, which only remain, are not half so valua le" (Essays 1. XIV, p. 126; G&G, p. 186).These first two kinds of rules make space for the third, the one we have been ca ling "cor ective" .These general rules are "the logic" of proba le reasoning and is required to achieve justified and relia le belief, on which a l valua le sciences are based.Cor ective general rules have therefore at least a threefold function: (1) they display a model of relia le belief formation and cor ection; (2) they can also cor ect judgment produced by the first kind of general rules (T 1.3.13.12;SBN 149-50); and (3) they make it possi le to return ir eflective judgment or belief to its cognitive sources and foundations and, thus, to identify false belief as such (see T 1.3.8.14; SBN 104-5).

General rules for causal reasoning
After developing his theory of philosophical probabilities and its dependency on causal inference, Hume outlined his famous set of Rules by hic to judge causes and effects (T 1.3.15;SBN 173).They are 8 criteria that a low us to distin-uish between a relation of constant conjunction that describes a causal nexus and an a parent causal relation.They should permit us to discern the cor ectness of inferences and beliefs based on causal reasoning.Furthermore, Hume affirms that the 8 rules are "a l the logic I think proper to employ in my reasoning" (T 1.3.15.11;SBN 175).They should rule our causal reasoning so that they resem le as much as possi le the Proof13 horizon.The general rules for causal reasoning are the natural conclusion of Hume's treatment of probabilities and proba le belief in the Treatise (and not the skeptical conclusion of T 1.4).This is not only because of the plain fact that, according to Hume, every reasoning concerning matters of fact relies on causal inference, but also and mostly, because despite this reliance "the relation of cause and effect has a l the op osite advantages" compared to reasoning based on "feign' d resem lance and conti uity" since "the objects it presents are fixt and unaltera le" (T 1.3.9.7; SBN 110).There are many other principles that enlivened our ideas similarly bringing us to believe "and command our assent beyond what experience wi l justify; which can proceed from nothing beside the resem lance between ideas and facts" (T 1.3.9.12; SBN 113), for example, credulity ("easy faith in the testimony of other") and education, which rest "almost on the same foundation of custom and repetition as our experience or reasoning from causes and effects" (T 1.3.9.19; SBN 117).As a result, accuracy in the determination of causes is needed to avoid the "inaccuracy" , which is "contrary to true philosophy" (T 1.3.9.19 footnote; SBN 117).The cor ective principles for causal reasoning, which are "a true ecies of reasoning and the strongest" (T 1.3.7 footnote), are a l the logic necessary in order to "rectify non philosophical probabilities into causal probabilities" thus achieving a true philosophy, upon which, according to Hume's project, depends the real possibility of reaching "a system of proofs, "14 or in other words: science.
As can be seen, this set of rules has a unique chara er.Since cause and effect is, properly eaking, the only relation of matters of fact that results in reasoning (T 1.3.2;SBN 73) -that is, drawing a conclusion from given premises, or the generation of new beliefs from given ones -the rules for judging cause and effect relations have to mediate this process of forming beliefs in order for that judgment to be an expression of an adequacy between the natural tendency of the mind and the "stu born nature" of the object of its judgment, both of which are necessary for developing science.They rectify judgments in so far as they evaluate if they are "subject to accidents or influenced by whim and private fancy" (Essays 1. XIV, p. 112; G&G, p. 175).
That is why I ultimately think Hume's account of corrective general rules, though terminologica ly akin to the use of modern logic, is different in nature in so far as it is the result of an inquiry into the deep nature of believing and judging.Those rules are not mere recommendations or simply an instrument for reaching certainty.They represent instead the structure of a cor ected natural faculty of reason; in other words: a standard.Precisely on this point rests the origin of their normativity 15 .I agree with Hearn's claim that "these rules come for Hume to occupy a different status [...] the function of the causal rules is to cor ect and stabilize the sentiment of belief which is generated by certain natural, causal factors" (1976, p. 65).It seems to be the case that cor ective general rules are normative rather than descriptive, that is, they are prescriptions about how we ought to form and cor ect states of belief.Now, this is sti l insufficient to prove that general rules have normative force, indeed, someone could insist that Hume is just describing the way we form more stu born and relia le beliefs, as he did with false beliefs.Hence the important issue here is to determine the extent to which the account of general rules belongs to this level of discourse, that is to say, whether the rules are meant to account for the doctrine of natural causes of belief or if there is something else to say about them.

General rules and the normative dimension of belief
The question concerning the normative dimension of belief runs into the intricate relationship between belief and truth.This relationship can be described as fo lows: to believe that p is to believe that p is true.Thus, p ima facie, a cor ect belief, that is, one that fulfi ls that pretension, is ultimately a true belief 16 .It fo lows, therefore, that a rational agent should believe p if and only if, there is enough evidence for the truth of p.The same would ap ly for the case of rules: if A, B and C are principles for forming true beliefs, it fo lows that in rea-soning (coming to a cor ect conclusion) we have to consider beliefs that are consistent with those rules to be more relia le than ones that are not. 17 Yet by itself, this relationship does not seem adequate to account for the source of normativity of belief.In fact, the assumption outlined in the above paragraph is exactly what needs to be proved here.For even if belief aims at truth, the fo lowing naturalist objection cannot be easily avoided: "To elevate this trivial fact to the status of a 'norm' is to transform an innocent platitude into a pompous falsehood.For there is nothing normative about believing: neither we believe with an eye fixed on the horizon of an ideal of truth nor we obey any prescription to believe the truth" (Engel, 2007, p. 179)  18 .In other words, the relation that our beliefs have to truth can be seen as a plain fact; it expresses the fact wherein one believes p, rather than a compulsory prescription about what to believe.
I would like to draw attention to the pro lem concerning the scope of normativity as a first step to present my response to the naturalist objection.Normativity, at least in a philosophical sense, is not mere necessity (logical or physical).Rather, it concerns what is not absolutely necessary and, accordingly, it would be pointless to attach in any way normativity to a plain fact as breathing or sunshine.The proper scope of normativity, in the sense I am intere ed in, is that of pra ice (see, for example, Stemmer, 2008, p. 32;Railton in Dancy, 2000, p. 4).If believing and breathing are not two different kinds of phenomenon I would agree that there is no point in ascribing normativity to the realm of belief.But I think that there is certainly a difference between them, in so far as belief is the result a typical kind of agency, namely, epistemic agency, which involves other typical components of the realm of normative facts, such as judgment, wi l, epistemic freedom and, in short, rationality.Now, the claim that believing is a subject of the wi l is -for good reason -controversial (see Owens, 2000), and though I cannot a dress this controversy here, I wi l instead, assume a position very close to McDowe l's (1998, p. 434f) and O 'Hagan's (2005, p. 45f), 15 Lyons rejects the claim that the general rules for causal reasoning are second order mental states, evident by their reflective character (T 1.3.13.11;SBN 149).Instead he holds that they are about objects (Lyons, 2001, p. 273, n. 13).I believe Lyons's claim is wrong since according to Hume causation cannot be objectively predicated on objects, but only as a projective function of the intellectual power of men.General rules are explicitly rules "to judge", thus, they refer to acts of the mind rather than to objects.Nevertheless, I agree with Lyons that they are reflective in the sense that they involve the idea of causation, which is an idea of reflection. 16In a deeper analysis this claim has to be qualified.Epistemic norms are in a sense standards of correctness of belief.Norms governing beliefs are nonetheless still related to their characteristic aim: truth.Nevertheless from the perspective of real epistemic agency, believing is not necessarily a matter of "all or nothing", but of degrees of rationality, certainty, correctness, evidence, assurance, confidence.This is also something that Hume has permanently in mind while dealing with probability (see for example T 1.3.7.2;SBN 130-31, 1.3.13.2;SNB 143).For Hume, different degrees of evidence constitute important epistemic distinctions, as is the case, for example, between probability and proof.For Hume there are some states of belief which are justified, and that justification is a function of their sources (if they are reliable or not), their stability and the dispositions of the epistemic agent to believe 17 For further developments on this issue see, in particular, Stemmer (2008, p. 77-79, 99).For the topic of normativity and epistemic norms, as well as norms of truth, see Lyon (2001), Owens (2003, p. 285-289) and Engel (2007, p. 182 ff). 18The objection is formulated by Engel, although he does not contend it.For a contrary position see O 'Hagan (2005, p. 44) and Stemmer (2008).Basically, these authors will sustain a "constitutive argument" in the sense that, as Stemmer for example put it, "it appears that the will-to-be-rational is an intrinsic goal of reasoning" ("Es kommt hinzu, dass das Rational-sein-wollen ein intrinsisches Telos des Überlegens ist", Stemmer, 2008, p. 60, the English translation is of my own).
in the sense that even if belief is not typica ly the result of our deciding what to believe, its occur ence is insepara le from, even constitutive of, the exercise of judgment, which cannot take place outside of the space of reasons.
Hence, in so far as freedom and rationality are involved in how we structure our doxastic lives, there are p ima facie good reasons to assume that there is also a normative dimension involved.Hume's theory penetrates this dimension by moving from a descriptive account of natural causes of belief formation to a deeper level, where reflective mechanisms of belief cor ection and formation reveal that belief is not a mere mechanical response 19 , but also a matter of rational deliberation.Cor ective general rules are mechanisms of reflective thinking, directed to judgment and, therefore, standards of rational thinking.Now, standards belong natura ly to the pra ice of reasoning itself, but they only become explicit if we assume a critical per ective and pay attention to what underlies the mere possibility of the pra ice.Reflection shows that it is inherent to epistemic agency to be normative, and this fact does not contradict in any measure the fact that real epistemic agents are most of the time insensi le to their being "responsive" to reasons.
Cor ective general rules are ca led upon to determine the standards of cor ection of causal judgment, and with it, the standards of cor ection of belief.A belief, accordingly, can be more or less adequate, depending on the degree of evidence and experience availa le.Furthermore, general rules help to determine the level of adequacy, since those experiences that can be captured (let themselves be explained) by general rules are what concerns science.It is important to keep this point in mind.Given a statement, there are certain conditions under which it should or should not be believed, that is, taken as true.Nonetheless, these conditions, according to my reading of Hume' s epistemology, are deeply related to a demarcation criterion, much more than a truth theory.A belief can report different levels of certainty, depending on its nearness to the poof horizon.Proofs are basica ly beliefs for which there has been no exception in experience, which is what general rules try to secure.We also know that Hume believes that those rules are "very easy in their invention, but extremely difficult in their application" (T 1.3.15.11;SBN 175).In my opinion, the account of general rules is far from being about how to ensure that one' s belief is true, as is the case with Descartes, for example; rather, it concerns the conditions under which a doxastic item can count as knowledge in Hume' s "liberalized" non-rationalistic sense, which somewhat "divorces questions of justification from questions of truth" (Lyons, 2001, p. 270) 20 .
A l this however, does not solve the pro lem of the normative force of epistemic rules, as uide of processes of belief formation and justification, i.e., why should we fo low these rules, and where does their normative constraint lie.

Normative force
General rules are cor ective, in so far as they state the standard of certain knowledge of causes.Furthermore, since generalization is a natural tendency of the mind, it fo lows that general rules are the consummation, the telos, of such a natural tendency.In other words, nobody expects or intends to fail at reasoning (in the broad, Humean sense of the word), even when the roots of that reasoning are not actua ly present to the mind.Everyone assumes that he or she reasons cor ectly -at least aspires to reason cor ectly -and thereby, that the beliefs he or she forms are cor ect, in the same sense that believing p necessarily involves believing its truth.Believing, as we l as reasoning, aim at truth as their intrinsic condition.Since causal reasoning (or causal inference or generalization) is inevita le, and since it natura ly involves the intention of truth, it fo lows that everyone must reason in accordance with general rules.In other words, "the authority of reasons is found within the pra ice of reasoning itself.We reasoners are bound by rational standards because to engage in reasoning just is to be accounta le to rational standards" (O 'Hagan, 2005, p. 43).Specifically, since a l matter of fact reasoning is a causal reasoning, one ought to pay heed to the 8 causal rules by which one may judge cause and effect in order to avoid false beliefs 21 .
I claim that believing and belief formation by proba le reasoning would constitute in themselves what Peter Stemmer has ca led a "normative situation" (Stemmer, 2008, §4), that is, a situation in which a normative ought-ness is implied, even if that situation is not epistemica ly present to the agent.The situation can be described as a hypothetical statement: "if we wish to achieve cor ect beliefs, we should reason according to some G rules" .If we wish the end, we are "normatively" required to act -to reason -in a certain way.The aim at truth of belief confi ures, so to eak, the normative situation.Now, for general rules to have normative force, it must be a condition that we do wish to have cor ect beliefs, so that they express not merely descriptive value, but also normative authority.This leads to the fo lowing question: why should we want to have true beliefs?Why should we be motivated to reason according to general rules?I think Hume also has an answer to this question, a question that is deeply connected with his naturalist conception of human reasoning: we need to reason cor ectly, because as agents we desire things, and in order to reach what we want, we need to identify the efficient means for obtaining them.Cor ect causal reasoning is a necessary condition for achieving the ends we desire; thus, being responsi le epistemic agents by reasoning according to basic epistemic norms is something we must do in order to satisfy our desire.This is also, according to Hume, the primary reason for why we engage in reasoning.Science and truth are secondary targets of reasoning, but certainly not of less importance; the first target, however, is instrumental.This conclusion shows, moreover, how deeply Hume' s epistemology is oriented to his Moral theory, and that an understanding of Hume' s account of reasoning and general rules sheds light on those passages in book 2 and 3 which at first sight seem to conflict with the "skeptical conclusion" of book 1 (see for example, T 2.3.3.6/7;SBN 416, T 2.3.10.1;SBN 449, T 3.1.1;SBN 458-9.In the secondary literature see e ecia ly Winters, 1979).
But since much of our causal reasoning is not conscious, most people do not manage to know that they have to fo low general rules in order to reason cor ectly, for this a ivity is already normative oriented.Even if they recognize that they have to reason carefu ly and according to general rules, they may choose not to do it, because they are influenced by other passions.But this is another matter, which has to do with rational deliberations and rational decisions.Lyons (2001, p. 270) has ar ued that "the normativity in Hume's epistemology" lies in that "the philosophical method derives its greater value from being a better means of satisfying curiosity (and keeping it satisfied) as we l as meeting other, daily, pragmatic ends" .I consider this to be cor ect, but instead of under-valuating it as a case of just "instrumentally-inspired" normativity, I believe, with Stemmer (2008, p. 33-44), that the source of normative ought-ness is -in most cases, at least -a relation of "necessary condition" to which is attached a wish/desire/want that actualizes the normative force of the condition 22 in as much as not fo lowing it, necessarily means not achieving what is desired.Here rests one of the most important of Hume's legacies: we are not a le to deeply understand (even) our most abstra ive cognitive processes without reference to the affective scope of human nature.In other words, there is an a ive exchange between the "sensitive" and the "cogitative" part of our nature.This exchange also takes place in the constitution of the nor ati e, as described by Fred Wilson: "Reflecting upon [...] experience we adopt goals that are attaina le and means that are efficient.Self-reflection leads to standards of pra ice that define the (cognitive) virtue of rationality; it leads to standards that are attaina le and efficient.In other words, self-reflection leads to a reasona le standard of rationality.Or at least, it does so if one is wise" (2008, p. 416).

Conclusion
The rudimentary tools with which Hume's philosophy and epistemology is equip ed make it, at first sight, look very unfit to be a contribution to the actual debate concerning epistemic normativity.However, his account of cor ective general rules provides a more or less persuasive account of the sources of the normative dimension in belief formation and cor ection, and it also provides a solution to the pro lem concerning motivation.Hume's ideas seem to be sup ortive of certain ways of a dressing the issue of normativity and epistemic agency, namely, constitutive strategies.Reflection thus uncovers the normative structure of belief itself, and belief 's aiming at truth implies a normative relation between epistemic norms and doxastic items.That is the way I think we have to understand Hume's claim that there are some "general rules, by which we ought to re ulate our judgment concerning causes and effects; and these rules are for 'd on the nature of ou understanding, and on ou expe ience of its operations in the judgments e for concerning objects" .But also the origin of their normative force is uncovered as lying ultimately in the inner psychology of human nature, deeply embe ded in the fa icity of life.For Hume, believing rightly, that is, believing what is "more proba ly" true, is determined significantly by utility, and by passions like the love of truth and cu iosity.Now, as he stresses, "the question is afte hat manne this utility and importance operate upon us?" (T 2.3.10.4;SBN 450).This issue is no longer a question for epistemology, but instead for the science of man as a whole.Natural dispositions, psychological mechanisms of the mind, and social and cultural constructions are ca led upon to answer it.
A l these nuances seem to confi ure the normative dimension of Hume's epistemology, which from the very natural ground of our "aiming at truth, " is oriented toward the improvement of the understanding and human chara er.Its significance is not diminished because of the distance of Hume's theory from the question of an objective truth, for as Owens (2003, p. 287) states: "Rational belief is rarely based on conclusive evidence" .Nonetheless, and from a certain technical per ective, Hume's epistemological ap roach is weak, because relevant questions, such as how much evidence, beyond his very general demarcation criteria, is necessary to ca l a belief "true" , are never settled.This lack of tidiness also shows that Hume's first concern was neither epistemological nor logical, but rather moral in the broad sense of the word.