Jusletter IT

Foundationalism requires that a beliefx is justified if and only if it is self-justified or not inferentially justified or if it is derived by a self-justified belief. According to foundationalism, the chain of beliefs ends up with a belief which, in turn, is fixed and does not need support. Thus,each belief is mutually related to other justified beliefs andno belief can be unlinked to the foundational belief¹ .

What beliefs are foundational? Basic beliefs can be self-evidently mathematical truths like the numbers which are the departure point for other statements, or geometry axioms such as the transitive axiom, or logical truths. In the light of this, and following Haack’s distinction, one may say that: «Some beliefs are basic; a basic belief is justified independently of the support of any other belief; basic beliefs are non-empirical in character »² . Another approach places more weight on the basic beliefs as self-justifying physical sensations. This is the way Empiricists argue: the «immediately given»³ is the observable through the senses. For instance, I do not need to give any foundation to the sensory belief that I smell burned milk: I simply smell it. I may say that I may wonder whether I lack appetitebecause of that smell but the basic physical experience (the smell of burned milk) remains foundationally reliable and fully justified. In the light of this, one may say that: «Some beliefs are basic; a basic belief is justified independently of the support of any other belief; basic beliefs are empirical in character ». The claim that «basic belief is justified independently of the support of any other belief » may fulfill two functions. In the first place, and following in the footsteps of Haack, it may serve as a replacement for «but by the subject’s experience », or «but because of a causal or law-like connection between the subject’s belief and the state of affairs which makes it true », or again «but in virtue of its content, its intrinsically self-justifying character »⁴ . Secondly, «a basic belief» may not require degrees of justification in the sense that it is «completely, conclusively justified independently of the support of other beliefs »; or it may demand degrees of justification in that it is «justified prima facie but defeasibly/to some degree but not completely, independently of the support of other beliefs ». The former version is called «strong foundationalism», while the latter «weak foundationalism»⁵ . In relation to basic beliefs, other supporters of foundationalism may refer to «facts about language-learning» or to the view according to which the issue of basic beliefs in justification do not rise at all⁶ .

Foundationalism is reminiscent of the Euclidian geometric paradigm considered as the most appropriate way of organizing a body of knowledge⁷ . The very long tradition of Euclidian model has framed the intellectual history of the Western science and philosophy, from Greek antiquity and has undisputedly dominated throughout the centuries. Irrespective of how the nature of foundationalism and the various specifications of foundationalism may include, one usual example for foundationalism concerns the rationalistic epistemological system of Descartes. In hisfifth reply to the Dutch theologian Caterus who disputed with him about the existence of a being more perfect than humans and about the possibility that humans could exist if such a being never existed, he said that uneducated folk miss to differentiate the concepts of causesof coming into existence from thatof staying in existence⁸ . In making this claim, he mentioned the following example: «the architect is a cause of a house, and a father is a cause of his child »⁹ . The architect and the father are only the causes of the house’ and the child’scoming in existence , but not the causes of theirstaying in existence . Furthermore, he claimed that the more perfect being – God – is the cause of thestaying into existence of the sun though the sun is the cause of thecoming into existence of the light it emits¹⁰ . Alike the building and the child whose existence remain «once the work is completed », the sun would not since the cause of its staying in existence (and not only in its coming to existence), that is God, must continuously work in keeping it in existence. In demonstrating what humans derive their existence from, Descartes stated the asymmetric interconnection between causes, of which one must be foundational: only if each cause is properly linked to the foundation is justified. He understands the existence of God as a necessary truth for all created things: without the ontological argument both the experience and Cogito itself would fall in pieces. The existence of God is the atom belief, from which he reconstructs the whole system of knowledge. In hisDiscourse on Method he develops a notion of justification which is heavily steeped in foundational terms. He argues that every belief must be rooted ultimately in indubitable knowledge. Given the power of reason, the thinker or the doubter can proceed from doubt to doubt up to the Cogito sum, theclara et distinct perceptio which is apart from receiving warrant from other beliefs. Descartes’ system of knowledge of external world is, via theology, founded on an indubitable super-structural premise. As an example for illustration, Descartes uses the metaphor of scientific method and the architect activity. I shall quote a passage from Descartes’Replies : «My method imitates that of the architect. When an architect wants to build a house which is stable on ground where there is a sandy topsoil over underlying rock, or clay, or some other firm base, he begins by digging out a set of trenches from which he removes the sand, and anything resting on or mixed in with the sand, so that he can lay his foundation on firm soil »¹¹ . It follows from the outlined analogy that knowledge is rooted in fixed foundations and that the beliefs which are not supported by basic foundations, as the doubtful beliefs, must be thrown like sand. Descartes’ entire plan, was in contrast with both skepticism view «who doubt only for the sake of doubting and pretend to be always uncertain » and the old way of doing science which «borrow principles from philosophy », «in order to find the rock or clay »¹² .

InContemporary Theories of Knowledge , Pollock and Cruz have classified the theories of knowledge in two main broad groups, Doxastic Theories and Nondoxastic Theories. In their view, foundations theories and coherence theories belong to the same area, in that the justifiability of a belief is a function exclusively of what beliefs one holds – of one’s «doxastic state» – independently of the circumstances under which the belief is held. The doxastic assumption consists in the fact that «all our information about the world is encapsulated in beliefs »¹³ , thus – alike from nondoxastic theories – no other considerations enter into the determination of whether a belief can be justified.

Similarly to foundationalism, coherentism shares the same idea that central to epistemology is the concept of system, in particular the concept of cognitive systematization which «facilitates understanding » and «provides the channels through which explanatory power can flow »¹⁴ . However, albeit the common peculiarity of ideal systematicity of knowledge, coherentism provides a substantial different answer to «what sort of system »¹⁵ better accommodates a proper cognitive systematization. Coherentism requires that a beliefx is justified in so far as it figures within a system of other beliefs. Thus,each belief is mutually related to other justified beliefs butno belief is linked to a foundational belief. Roughly speaking, coherentists do considereach belief as relationally supporting one other butno belief is to be taken as super-structural or as especially privileged. This suggests that no belief is given as ultimate foundation of a system rather that all beliefs are in principle revisable in their capacity of being confirmed, modified or rejected. Knowledge does not arise from the assumption of linearity among inferences of beliefs some of which occupy a privileged position, but form their capacity of receiving warrant from other beliefs. In this view, justification is uniquely a matter of relations among beliefs: justification occurs if and only if a beliefx coheres or fits into a whole¹⁶ . How do the beliefs fit into a whole? Haack, for example, draws a distinction between an «egalitarian form of coherentism» according to which all the beliefs in a coherent set are exactly on a par with respect to their justification and the «inegalitarian form of coherentism». The former version can be characterized thus: «A belief is justified if it belongs to a coherent set of beliefs, no belief having a distinguished epistemic status and no belief having a distinguished place within a coherent set ». According to the latter: «A belief is justified if it belongs to a coherent set of beliefs, some belief having a distinguished initial status, and justification depending on weighted mutual support »¹⁷ .

Coherence is not equitable with logical consistency: they are possibly related concepts which do not necessarily imply each other. Suppose that I believe a scientific model that says that for snow to fall the temperature is to be of 40 degrees; suppose that my observation clearly shows the opposite as I am driving a jeep in the middle of the Erg desert where the temperature reached is exactly 40 degrees. The former belief is beyond a doubtinconsistent with my observation: either the former or the latter must be rejected. Nevertheless, the belief in the crazy scientific assumption may beconsistent with the belief that going to Grieg’sAnitra ’s dance performance would heal my headache. Thus, one thing is to say that tough unrelated, beliefs may be consistent to each other; another thing is to say that the beliefs are mutually supporting in such a way to yield a working scheme. In this case, the belief that it may snow at the temperature of 40 degrees isconsistent with the belief that going to the theatre would help my headache, but it isincoherent with the whole of other scientifically related beliefs regarding physics, studies of climate, geography, and so on. Bonjour observed that «a coherent system must be consistent, but a consistent system need not to be very coherent»¹⁸ . It should be particularly emphasized that consistency is not the same as coherence¹⁹ and that coherence requires more than consistency as it is a quantity concept of less or more degree: the larger (or the less self-closed) is the system where we place our beliefs in, the more coherent it is in comparison with any currently alternatives, the truer the beliefs are.

From early the Fifties Quine’s opposition to the analytic-synthetic distinction arose to defend the coherentism program and to crucially reform epistemology. In his well known «Two dogmas of Empiricism », Quine claimed that the strict distinction between the sentences whose value-truth is found only on the meaning of its words (analytical sentences) and the sentences whose value-truth found on its meaning as well as on facts (synthetic sentences), could not properly accommodate the task of justification of knowledge. «That there is such a distinction to be drawn at all – Quine maintained– is an unempirical dogma of empiricists, a metaphysical article of faith »²⁰ . Furthermore, he also rejected the second dogma of empiricism according to which every meaningful sentence is structurally made of experience terms: roughly said, he denied one of the strongest empiricists assumption according to which every sentence may be confirmed or disconfirmed by an empirical verification. His enterprise was clearly aimed at deflating the positivist and metaphysical pretentious picture of rationality as fundamentally aprioristic to science itself. The new architecture of Empiricism would capture how, given only the evidence of our senses data, we humans do not deduce but construct our scientific theory of the world.

Keyword for the coherence model of justification is holism (from óλoς «the whole», «the entire»). By rejecting the linear conception of inferential justification, the Quinean holistic principle to which science and epistemology obeys, offers an account of the interconnection among various elements of a web system: each belief (or each sentence) is justified as far as it is not immune (or relatively immune) to revision in reference with a larger theory²¹ . The relevant consequence of this perspective is that all beliefs (and statements) are not in need of being ultimately justified: beliefs are not of two main kinds, foundational and derivable, but the justification of beliefs occurs in virtue of the mutual support. This, in turn means that all beliefs are in the sameboat : like sailors who must rebuild their ship on the open sea and reconstruct it by only adapting the available materials in the face of changing circumstances as best they can, scientists do. The boat and the materials on board is all what sailors have: for the sailor there is no dry dock when repairing the boat, neither privileged nor best materials or solid ground when dismantelling the boat. A dry dock is the same happening to scientists: no solid foundation from which to work on and build up a conclusive knowledge, only there is no best material for secure knowledge. This is Neurath’s notorious metaphor²² which discretely shows the perplexity for the positivistic foundational canons of knowledge and mitigates the hierarchical approach to justification²³ .

Davidson’s works in the philosophy of language show that in order to interpret a belief we must consider our own system of beliefs as wholly true or in agreement with us. A feature of Davidson’s discussion of radical interpretation is that the achievement of a merge of theory of beliefs and a theory of meaning is through the application of the principle of charity. He asserts that «we cannot intelligibly attribute any propositional attitude to an agent except within the framework of a viable theory of his beliefs, desires, intentions, and decisions » and that particular beliefs make sense: «only as they cohere with other beliefs, with preferences, with intentions, hopes, fears, expectations, and the rest (…) the content of a propositional attitude derives from its place in the pattern (…) »²⁴ .

The coherentist method applies also in philosophy and theory of science. Goodman discarded the «notion of a necessary connection of idea, or of an absolutely analytic statement »²⁵ , by bringing one of the coherentist arguments concerning the justification of the principles of deductive and inductive inference to central stage. How is a deduction justified? Evidently, by conforming it to the general rules of deductive inference, independently of its truth or falsehood. To justify a deductive conclusion it is neither important to be aware of the facts the deductive conclusions pertains to, nor to acquire knowledge of what justifies the rules of logical inference to which the deductive argument conforms to. The old problem of what the justification would constitute, is not peculiar to deduction in so far it analogously applies to induction: Goodman kept the question whether the validity of induction relies upon the conformity to valid rules of induction, or else. What if there existed more than one valid rule, or more valid rules were equally valid? Where do these rules follow from? Certain philosophical positions as foundationalism, might, however, have discussed self-evident axioms, or the very nature of human mind²⁶ . But if one goes more closely into the old problem of justification, one will see that a better understanding must be gained. Principles of deductive and/or inductive inference are justified if and only if they conform to accepted practice. A justification purports to be valid if the general rules and the inferences are brought into agreement with each other. Goodman formulates this aim as follows: «A rule is amended if it yields an inference we are unwilling to accept; an inference is rejected if it violets a rule we are unwilling to amend »²⁷ . Given the complexity of the matter, the application of the notion of «mutual adjustment» is not delimited only with reference to the justification for deduction and induction, but it covers a wider area. Goodman brings into focus that «the interplay between rules of induction and particular inductive references is simply an instance of this characteristic dual adjustment between definition and usage »²⁸ . Plainly, the coherentist concept of «mutual adjustment» is developed outside of the theory of science and it is applied to the theory of language, with particular regard to definitions: «a definition may modify as well as extend ordinary usage »²⁹ . In so far as we accept to extend the usage of a definition, we may accept to extend the term «valid induction» to other inductive inferences not usually considered as such³⁰ .

Notoriously, a successful invocation of coherentism is displayed by Rawls’ theory of justice. Goodman’s notion of «mutual adjustments» is mainspring to Rawls’ reflective equilibrium method³¹ . In his inquiry into the role of justice and fairness, Rawls took distance from the traditional foundational doctrines such as classic utilitarianism and proposed an alternative in line with the classic tradition of social contract as found in Locke, Rousseau, and Kant³² . In a purely original position, characterized by equal position, social status and opportunities, and the like, individuals as moral persons are symmetrically «rational and mutually disinterested »³³ . In such a initial status quo, men set up the first principles of justice as fairness, and accordingly, a constitution and a legislature to enact laws. Rawls’ theory of justice meets his theory of rational choice: how would the chosen principles match the individuals’ considered conviction of justice? How to accommodate our provisional fixed convictions, for instance that the racism or to use violence against another person are unfair, with the principles in such a way to impede discrepancies? «By going back and forth » the principles and the considered judgments, we shall revise, modify or confirm the judgments, so Rawls maintained. The relationship between on the one hand the considered judgments and the matter of justification and, on the other hand the task of developing a substantive theory of justice comes out clearly in the following quotation: «(…) justification rests upon the entire conception and how it fits in with and organizes our considered judgments in reflective equilibrium (…) justification is a matter of the mutual support of many considerations, of everything fitting together into one coherent view »³⁴ .

Regularity in attribution the label foundationalist or coherentist to any philosophers (epistemologists, moral philosophers, legal philosophers) is too hard and even if possible, uninteresting. An idealized division line might drop irrelevance and lead to unfashionable abstractions (with the exception, for example, of someone who labels himself coherence theorists, like Davidson³⁵ ). This is due to the fact that theorists may hold approximately similar views and be very close in spirit though not sharing the same conclusions, and vice-versa. The terms «foundationalism» and «coherentism», like many other ism-terms used in literature, may lead to problematic categorizations which would arouse expectations of a minimum of unity. An example is provided by one of the most brilliant contemporary epistemologists, Susan Haack, who has persuasively articulated her own theory on the assumption that no fruitful discussion may sparkle under strict dichotomies. Her project takes shape on a certain hostility concerning untenable epistemological dualisms, given that the whole epistemological enterprise cannot be reduced to familiar rivalry of foundationalism versus coherentism. They both do not exhaust the field of knowledge: categorizing leads to problematisations and leaves room for imprecisions and lacunae. By virtue of this, Haack favours an impressively original intermediate theory between the traditional rivals, neither foundationalism nor coherentism: foundherentism. She says it may be approximately characterized thus: «A subject’s experience is relevant to the justification of his empirical beliefs, but there need be no privileged class of empirical beliefs justified exclusively by the support of experience, independently of the support of other beliefs »³⁶ . By the perspective of foundherentism she means a theory which allows both the relevance of experience to justification and the importance of mutual support. She restricts her inquiry to empirical basic beliefs (thus disregarding the non-empirical beliefs) and acknowledges the supportive mutuality among a person’s beliefs. It turns out that like foundationalism, beliefs are not essentially one-directional of evidential support and, alike coherentism beliefs are not depending solely on logical or quasi-logical relation among beliefs. To put this in general terms: «Justification is not exclusively one-directional, but involves pervasive relations of mutual support »³⁷ . The fact that the phenomenon of justification is nevertheless not solely found in foundationalism is explained, inter alia, because there is no bodies of beliefs that have to be seen to be justified solely by the subject’s personal experience. Justification is construed depends also on the pervasive mutual support among a person’s beliefs. Seen from one side, justification relies on the ‘evidence of the senses’: Haack maintains that we human beings see and interacts, through the means of our senses, with the things surrounding us. How justified a person A is in believing that-p? The answer to the question depends on why he believes that-p, in his S-beliefs and experiences, i.e. the causesoperative at the time in question³⁸ . To recognize such causes is to refer to a balance of factors which inclined him towards believing that-p or against believing that-p. At last, to refer to a balance of factors is to draw a distinction between which are states of the person concerned form those which are not: the components within the causal nexus of A’s S-belief that-p may be, to varying degrees, seen as perceptual states, introspective states, memory states, etc. Yet, the justification of her belief is not one-directional conceived thus it aspires to be gradational and not linearly ordered as if it is when one determines the soundness of a mathematical proof. The knowledge enterprise is imaginatively comparable to the reasonableness of determining the entries in a crossword puzzle. Haack formulates her aim as follows: «the clues are the analogues of the subject’s experiential evidence; already filled-in entries, but the entries are, in variable degree, interdependent; these are the analogues of the asymmetries already noted between experiential evidence and reasons »³⁹ . The reasonableness («how good») of a certain entry depends on how much support is given to this entry by the clue and how much support is given by the intersecting entries already filled in; the reasonableness in believing in the correctness of the other entries, independently of the entry in question; the amount of intersecting entries filled in. Said otherwise, «how good» is A’s evidence referred to believing that-p depends on howfavorable (supportiveness), howsecure (independent security) and howcomprehensive (comprehensiveness) it is with respect to p.

The concept of justification laid down as a pure summary of «foundationalism» or «coherentism» lets it precisely remain open whether it is un/controversial to incorporate in one theory elements from both. Haack’s project fulfills people’s expectation that philosophers should create something new, and this is so true that her arguments provide new ground rules for the debate in epistemology. Following Haack, what I have tried to demonstrate is that there is plenty of room for combination that forms a reconstruction of not-locked models of justification and how misleading might be to squeeze any author into a pre-defined suit. Nevertheless, this question falls outside the main thread of the present discussion. Thepurpose of this work is actually to practising a distinction between foundationalism and coherentism – with particular reference to coherence theories of law – and to pointing out a common pattern in thought, thus cutting across considerable differences. Practising a dualism may have a value in its capability of providing a degree of awareness of discrepancies between problematic perspectives.

I have grouped the further discussion as follows:

• Foundationalism and natural law theories
  
• Coherence and coherence theories of law
  
• Coherence and the monstrum
  
• Coherence, the monstrum and the law
  

Foundationalism is the distinctive model of reasoning which has particularly gained consensus from supporters of natural law theories and more generally, from ontological accounts of law. No doctrine has been so deeply inspired by the idea ofnature counting as the foundation for Law. Since thousand years, a recurring approach in natural law theories had been to confirm or disconfirm whatexists with what itought to exist . According to natural theories, there is a foundational or super-structural belief which provides the best model determined for the reality: ideas such as human dignity, personal freedom, equality, principles for the governance of a state such as the separation of powers, the administration of justice, the common welfare and so forth⁴⁰ , move in a linear fashion straightly to a super-structural belief,the nature . It seems to be a general feature that Natural Law theories uphold the conviction that all ought to be-content is derivable from the concept of nature, and that heterogeneous contents fall under the label «nature». Platonic doctrine of ideas was an aprioristic doctrine and the backbone of his concept of natural law. In contrast with the relativism of sophism, he grounded knowledge on eternal truths which provided for the foundation of natural law. Aristotle’s inquiry – alike from that of Plato – incisively encompassed the connection between law and human nature, in terms of teleological metaphysics. In his view,Idea andPhysis interplay in that the idea is not outside the object but immanent to it and the matter tends towards the final form within a rationally teleological movement. He considered the nature as the end result of the movement, thus the natural law is what comports with our nature. Said otherwise, Aristotle’s conception of natural law is found in the formula according to which whatis conformed to nature corresponds to whatought to be in the best possible state. One implication is that if one wishes to know what is to be «conformed to nature» or «contrary to nature», then one must acknowledge what is «the best possible state» of a thing. The rationalistic approach was the work of medieval and modern minds. Saint Augustine inherited the Platonic doctrine of ideas and described God as the foundational principle, thus he gathered up the threads of classical antiquity and Christian tradition. On the wake of him, Scotus maintained that the first table of the Decalogue consisted in laws in the strict sense (existing independently of any will), while the second table of the Decalogue consisted in laws «very consonant with» it (thus, depending on God’s will). More radically, Ockham held that all natural laws are inferred from God’s good will, therefore all natural laws depends upon a self-evident principle and can be derived by deductive reasoning. Aquinas followed the Aristotelian teleological concept of nature depending on reason and enriched it by the fruitful contribution of stoicism doctrine which maintained the existence of ajus naturalis as emanation of thelex aeterna . In modern age, the concept ofrecta ratio which was fundamental for Scholastic philosophy’ and the Catholic Theology’s systems of knowledge, turned on a dime and moved from the concept ofrecta ratio . A new natural science-oriented rationalism took shape: Descartes’ synthetic-analytic method was grounded on the human concept-based ofclara et distincta perceptio, and provided the basis for any certain knowledge. It became the model of a new doctrine of natural law. The consequence was that from the XVII century the Aristotelian-Thomistic view of «human nature» encountered the analysis of the natural science and led to a different comprehension of nature. Human nature was considered as not having in itself the source of its order, indeed as an empirical organism, thus interpretable in terms of casuistic laws. This was, with respect to a variety of colors, the central view held by philosophers in the centuries ahead up to the beginning of XX century. From XVII century, a new attempt of systematically deriving legal material contents as generally valid from the human nature aroused⁴¹ . Foundationalism derived moral standards and political values from the very existence of human beings considered as social animals, as well as theorems are derived from a finite class of axioms in Euclidean geometry. The works of Grotius and Pufendorf are perhaps the best examples of the close connection between the human nature, physically and rationally conceived, and natural law. In XVIII century, human nature and law encountered in the Kantian idea of natural law as law of reason: reason became the source of normativity. Kantians, new-Hegelians, Utilitarians assumed the very nature of human beings as unquestionable truth and encouraged the picture of a social and political order in conformity with universal principles of objective morality. The state of nature gives origin to a universal order.

The main emphasis in the examples I have given an account of here is on many forms and degrees of foundationalism in the various schools of natural law. However, under the banner of natural law, the general phenomenon of deductive system remains the same. One worry that may arise in appealing to such external and mysterious concept of nature is that a special belief has some sort of privileged status which is immune to revision. Therefore, this model of reasoning is foundational in the sense that the concept of nature constitutes a fixed pattern, which is the product of an axiological decision and the basis on which all the system relies upon. Typical of foundationalist views, the apparent solution of one case is depending upon another, thus generating a chain of arguments moving all the way up to some general universal law of nature. There is not way-out within the chain of arguments: thepetitio principii consists in giving rise to an apparent solution insofar the justification of one concept barely gives rise to another, whose solution merely brings back to the first one. As previously pointed out, foundationalism appears to be caught into a vicious circle: by defining what is «good» by the means of «what is conformed to nature», one ties the concept of nature to a value. What one wishes to be defined as «good» consists in what is agreed to be «conformed to nature» and, consequently the «good» is deduced from what is «conformed to nature». The definition of «nature» lies on another, which is in fact depending on the first one.

Against the Euclidian linear fashion-model of justification and the positivistic model of justification, a coherentist new orientation was brought it to prominence, both in legal theory and interpretation, at least from the Seventieth. As main competitor of foundationalism, coherentism embodied a common pain response due to the agonizing deconstruction of legal argumentation pursued by Legal Realism (for example, in Scandinavia, with Hägerström and Ross) and Positivist schools. The straight application of coherentism in legal theory identified the quest for justification of law as morally possible, thus inversing the claim that moral values do not matter in legal argumentation. In what follows and, for the sake of brevity I use the word «legal coherence» with particular reference to Dworkin’s account of law as integrity as it captures a common tendency of coherentism vogue in the philosophy of law⁴² .

I shall point out some common features:

• From a structural point of view, a legal system is characterized by the coherence of its elements: no element is ultimately foundational in that only exist approximately fixed points, revisable in principle, exist. As already appeared, according to the holistic conception, the completeness of the legal system cannot be irrespective of the total mutuality of each element. In this spirit, Bonjour distinguishes two levels at which justification occurs. On the one hand, the justification of a particular belief (or a small set of beliefs) can be raised: at thislocal level, justification appears linear in the sense that «a given justification belief is justified explicitly by citing other premise-beliefs from which it may be inferred »⁴³ . At this level the justification of the whole context to which the particular belief is a member, is not at issue. On the other hand, justification can challenge the overall context of a cognitive system. At theglobal level, justification does not appear linear: the relation among the beliefs consists in their mutual support. This means that the justification of each particular belief rests on the «overall system and its coherence »⁴⁴ . The «interlacing network »⁴⁵ of linkages has not hierarchical structure, albeit it can be the result of deductive inferences. The overall system does not proceed along a linear mode, rather it can also portrait the image of a cycle. Following in Rescher’s words: «what determines correctness is the matter of over-all fit, through which every element of the whole interlocks with some others »⁴⁶ .
  
• Thus, absence of any ultimate connection whatsoever does not follow from the absence of certain univocal coherence. Indeed, with regard to law the univocal coherence is theratio essendi of a legal system. In the light of this, and given the symmetry between the completeness and the coherence, one can see law as whole axiologically cohesive. The coherence occurs at the level of the principles, in so far norms are linguistic formulations which speak directly of principles⁴⁷ . Although coherentism does not draw the picture of a pyramidal system of norms and principles, principles are required to be the ultimate reasons of the so structured legal system.
  
• By rejecting one of the classic assumptions of legal positivism, according to which a legal system is conceived as a whole of rules identifiable by means of a foundational master rule (i.e. a Grundnorm, a rule of recognition), one may affirm that in view of coherentism:
  
• A legal system is characterized by a merge of rules and principles whilst the validity of norms relies on formal reasons, the principles are valid as rightness reasons. In a different way, one ought to distinguish within the normative standards for law, between formal reasons and substantial reasons, or else between an explicative and a justificatory dimension within legal reasoning⁴⁸ .
  
• The whole of principles provide the genuinely justificatory reasons for a legal system, i.e. the moral foundations⁴⁹ (this specification makes the sources thesis play no role. The coherentism generally rejects the positivistic and realistic view according to which both the existence and the content of law depend upon its factual sources – the facts of promulgation and/or the application – thus not upon moral arguments);
  
• Alike from rules which may be hierarchically ordinate, principles fit into a straight axiological structure (the dogmatic model of justification typical of natural law doctrines and legal positivism should therefore be replaced by an holistically interpretative view).
  
• In such a perspective, for any system to be complete and coherent, it entails weighing and balancing. The same goes for a legal system: a common view pointed out by coherentism is that the legal practice necessarily requires balancing of principles and underlying legal reasons⁵⁰ . This suggests that principles are, in principle, ordinarily ranked, given the capability for each legal system of embodying values. First, in so far as principles are not merely summaries of existing positive legal materials but have law-producing force and case-resolution capability, law is complete and coherent. Given its completeness and coherence, law advocates the means through which it makes it possible to weigh and balance principles with principles and positive norms with the underlying principles. The leading thought of coherentism is that certain principles – as the principle of justice as fairness in Rawls’ well-known formulation – are ideals of a fair legal-political structure. For instance, Dworkin captures this claim in «the virtues of fairness, justice and procedural due process »⁵¹ . Fairness is a value concerning the distribution of power in the right way, while justice is a guideline for making decisions in the right way; procedural due process is «a matter of the right procedures for judging whether some citizens has violated laws laid down by the political procedures »⁵² . In a related vein, Bermejo reduces the complexity of changing legal systems to two main coordinates such as the coherence with the past, which corresponds to Dworkin’s sense of fairness, and the coherence with the future, comparable to Dworkin’s concept of justice. Thus, in the efforts to carry the existing positively norms to higher levels of abstraction, coherentism arranges a list of generalized principles to which a legal system rests on.
  
• The juncture of the clear-cut division between rules and principles and the pervasive weighing and balancing evaluation which are necessarily to be brought into decision, gives a special account of conflict resolutions strategies. One of the ways legal positivism intended completeness in law was the dogma of strictly deductive reasoning and the occurrence of only conflicts among norms⁵³ . With respect to how to solve norms antinomies, positivism usually argues for the application of the three canonical criteria:lex superior derogat inferiori (a superior norm suppresses inferior norm),lex posterior derogat priori (a later norm abrogates an earlier norm),lex specialis derogat generali (the more special rule prevails over the general one). A characteristic result of the coherentist perspective is that the application of certain criteria is not aimed at reconciling the collusion between principles and between norms and principles,all-things-considered . Antinomies and tensions harassing within legal reasoning occur only at the level of norms. In short, the truly aspect of the coherentist enterprise is that either the conflicts do not occur in the sense that the whole legal system relies on an axiological eurythmics. The hierarchy is distrust as equivocal, given that no conflict occurs beneath the surface⁵⁴ . Or, as a variant on the foregoing irenic argument, conflicts fail to disclose them or simply do not exist, for they figure as only theoretically admissible cases. The incompatibilities pertinent to a given legal order are radically reduced by the fact that principles are adjusted in such a way to yield a hierarchical working scheme hinged on the governing rule that a superior principle always suppresses the inferior one.
  
• Common to the instances mentioned is that any interpreter who ventures into the law enterprise must presume that law is a cohesive whole of principles. What lawis corresponds to what lawought to be in the sense that it offers a presupposed meaning aspect, in that the interpreter is aware from the start that the interpretation he applies has a value beyond.
  
• Although the interpretation is bound to remain a very difficult endeavortout-court , the interpreter construes by being aware that – in Dworkin’s words – «some set of reasonably plausible principles can be found, for each general department of law he must enforce, that fit well enough to count as an eligible interpretation of it »⁵⁵ . As already been apparent, coherentism requires the acceptance of principles whose justification is a matter of mutual support and fit «together in a coherent view »⁵⁶ . Self-founded principles or self-evident axioms are not necessary imposed conditions. In his anti-foundationalist account of political liberalism, Rawls on the one hand aimed to provide an account of justified choice that does not rely on self-evident truths but rests upon a process of modulation between initial convictions about what is good and bad and considered judgments in reflective equilibrium. On the other hand, he argues that «there is one interpretation of the initial situation (the original position) which best expresses the conditions that are widely thought reasonable to impose on the choice of principles »⁵⁷ . More effectively, the concept of reflexive equilibrium is transportable within legal interpretation: Dworkin argues that legal materials a jurist or a judge must take into account are interpreted in the light of competing political or moral theories matching pre-theoretical intuitions. As in morality, the test of this theory is that is should determine the one interpretation which make best sense of our practice of law⁵⁸ . The best interpretation,all-things-considered , is the product of the best choice, and the best choice gives rise to theone right answer : the judges who accepts to be bound to the ideal of integrity will figure out which is «the best constructive interpretation of the political structure and legal doctrine of (the) their community »⁵⁹ . The notorious figure of the idealized judge drawn by Dworkin, Hercules, embodies the coherentist model of justification in both interpretation and adjudication of law. Hercules is the one who, with backward-forward eyes, removes incompatibilities and oppositions, finds the dimension of fit and converts the accumulation of plural theories in one exclusive sphere. How much of a divine portent, an unprincipled man of genius, amonstrum is there involved with him?
  

I take my point of departure in objections regarding Hercules’ rhetorical device which might possibly be taken as a support for the coherentist view. See in isolation, Hercules – with reference to the legal materials – tests «any part of the great network of political structure and decisions of his community »⁶⁰ in such a way to form a coherent theory justifying the network as a coherent whole. Law as integrity requires the figure of a supernatural judge whose convictions about the cardinal virtues of justice and fairness represent what is the morally «best constructive interpretation of the political structure and legal doctrine of (her) community »⁶¹ . Even though competing legal theories say something different about ideas of justice and fairness, and judges may embrace «different higher-level opinions about the best resolution of conflicts between these two political ideals »⁶² , the considerable greater certainty is that judgesvolitionally choose. What is crucial is not whether the judge chooses some options and whether some options are tenable or untenable, but the fact that he canvolitionally choose the best option all-things-considered. The crux of the matter is, therefore, that the reasons for choosing one or another option are fraught with her volitional exercise: the volitional determinants compel and drive the judge’s area of convictions and beliefs. This assumption exposes coherentism to various difficulties and some objections may be raised and investigated as follows:

• The state of acting: is there always any good reason?
  

• The state of acting: the akrasia
  

The information immediately available to the agent are not necessarily the only ones. Current information may only lead to «provisional conclusions» which can be confirmed, modified or rejected on the basis of further information⁷¹ . To put the information in a coherent framework we must acknowledge that «each one of us is exposed to new information, often challenging the information we already have »⁷² . In this regard, Pollock identifies some examples which clearly proof that some kinds of reasoning are not-deductively valid, though they provide justification on their conclusions. I can enumerate a few of them⁷³ :

1. Perception: perception is defeasible. For instance, I may believe that a freesia is white, having perceived its color. Such a perception is defeasible when further information come to light: for instance, that the sky makes a color contrast which modifies my vision;
   
   
2. Induction: induction is known to be not-deductive reasoning. For instance, having always seen white freesias gives me a reason to believe that all freesias are white. This conclusion can be defeated as soon as I discover a pinkish freesia;
   
   
3. Probabilistic reasoning: this kind of reasoning is tightly linked to induction;
   
   
4. Temporal projections: the assumption of stability of time is also defeasible. For instance, my yesterday belief that there were fields full of freesias gives me a reason to actually believe that they are still there today. This conclusion can be defeated if I discover that all freesias were picked up early this morning.

Collisions about information may generate incompatibility (for instance, between two mental states with reference to the same object): arebutting collision occurs when:

• M is a reason to adopt Q,
  
• M* is a reason to adopt Q*, and
  
• Q is incompatible with Q*
  

Reason collision can also occur when one has a reason to believe that under certain circumstances a reason does not provide support to the conclusion. Schematically, anundercutting collision occurs when:

• M is a reason for adopting Q,
  
• M* is a reason for believing that M does not support Q.
  

Said otherwise, the undercutting reason gives a reason to believe that under certain circumstances, the truth of the premises does not warrant the truth of the conclusion.

• The state of acting and moral luck
  

Under the name of Hercules, Dworkin gave expression to a family resemblance of concepts as a form of unity. The judge’s preferences between eligible interpretations are replaced by her own moral and political convictions which «will reflect not only her opinion about justice and fairness but his higher-order convictions about how these ideals should be compromised when they compete »⁸¹ . Accordingly, the judge in order to decide which resolution of the legal conflict is the best, personifies the community and speak with one voice, the voice of moral authority. In doing so, he ignores his intra-personal conflicts⁸² as well as the academic boundaries and the compartmentalization of law, and merges the benefits of private and public reasons. The successfulone right answer , removed from considerations ofakrasia , epistemic errors, lack of information, or indeed luck, regret and frustration is presented as a career open not only to a talented judge but to a prodigious monstrum who stands up in defense of the world’s fairness and justice.

In this there may lie dangers. Specifically moral considerations make themselves felt in the form of considerations external to the decision making procedure. A desire for a just outcome encourages the development of comprehensive doctrines spoken in third person. A constantly recurring problem in the thinking outlined is that the voice of the community gives rise to themorally best interpretation of the political culture. If one acknowledges that advancing the considerations of moral certainty as a necessary condition for any meaningful interpretative assertionall-things-considered , then coherentism finds anchorage in foundationalism. Most of epistemic coherentists use the word «coherence» with special reference to «fitting together»⁸³ : the justification lies in the capability for each belief of mutually reinforcing each other. The only one right answer does not make sense to the image of acontinuum among the web-relation elements of a system and undermines the incommensurability among values by providing a systematic resolution of value. Behind this and to rescue the coherentism from the risk of grading off into the discourse of foundationalism, there appears to lie a disagreement about how coherence is to relate to in/commensuarability of values.

The point of this art is about weighing and balancing, given that they are the central factors in evaluation of alternatives. Does the evaluation of alternatives presuppose a unitary concept of weighing and balancing? Does the weighing and balancing presuppose a unitary concept of what is to be weighed and balanced? Does what is to be weighed and balanced possess comparable properties? A question is whether the weighing and balancing are unitary concepts displaying a common unit of measurement, yet another question is whether the balancing presupposes that what is to be balanced is unitary or homogeneously includable within a canonical list of abstract entities⁸⁴ . The discussion is especially intrigued when axiological aspects such as «utility», «social utility», «welfare», «social welfare», «justice», «social justice» and so on are invoked. On the one hand, a balance requires comparable quantities, and on the other hand comparable qualities. Ethics and Law are more like general theory of knowledge than Physics: reasoning in Ethics and Law do not limit themselves to claims of quantifiable physical objects measurable in dimension, subjected to techniques and described in mathematical terms. Practical decisions involve quality factors, but they often depend on a complex set of different properties whose combination make the factors widely incomparable. Factors on weight recall the weight of weight factors in such a way to implicate a further involvement of other relevant factors which can run parallel or pull in different direction⁸⁵ . Our decisions make a special claims on us: we oscillate from agent-centered views to objective views and this division is – in Nagel’s account – what renders values formally different and deductive methods fragmented. Values display formal differences in that they differ in the kinds of sources: «not all values represent the pursuit of some single good in a variety of settings »⁸⁶ . Insofar as the shape of our lives is comparable, the values which enter into our private and public commitments differ. For instance, the origin of utilitarian values is pretty distant from that of the Aristotle’s eudemonia as the latter do not take the quantitative aspect of the value achievements into consideration. Different values, basically irreducible to the same foundations, may provide guidance to alternative courses of action without the reference to common measurable factors. Nagel argues that «the legitimate grounds of decision are extremely various and understood to different degrees »⁸⁷ . The formally different origins about values fragments the coherent unitary of a system of priorities among values.

This argument has potential force against those who contend that legal systems are cohesive and complete and base upon the presupposition of top-down deductive reasoning working from the more general case to the more specific one. The imaginary figure of Hercules portrays the organizing principle of the coherentist system: the system generates an appropriate alignment of all reasons, moral and legal, and produces the warrant of irenic conflicts resolution. All the dilemmas brought into the system are reduced to powder, given that no conflict between law and morality is admissible as much as coherence (integrity) «fuses citizens’ moral and political lives: it asks the good citizen (…) to interpret the common scheme of justice » and «infuses political and private occasions each with the spirit of the other to the benefit of both »⁸⁸ . Expectations of such a kind are causes in themselves of violation of Hume’s guillotine:is the legal system coherently complete orought it to be so? The bizarre creating power of moral certainty is what lies under the doctrines inspired to the model of coherentism in the various forms of a foundational moral fact or a foundational argumentative fact. In view of the factual diversity of this state of affairs, it is difficult to tell the difference between coherentism and foundationalism as much as it is difficult to tell the twins apart⁸⁹ .

For purpose of this inquiry, this argument confirms the coherentism central tendency in ideal revival, the reverential dogma of objectivity, a dogma on which every system is founded. Coherence theories of law are combined with the idea that each legal system is cohesive in virtue of top principles and that the whole system deductively instantiates ideals of justice. On the one hand, coherentism constraints itself to the theoretical frameworks of ethical cognitivism or to externalism about reason⁹⁰ , thus it contents itself with sharing the same claims as the supporters of rationalistic natural law doctrines. On the other hand, it displays a unified spirit in what it applies sound dogmas variably adopted, from time to time, by different accounts of law inspired to systematic rationalism: the legal certainty; the moral certainty; the will, the system uniformity, completeness, cohesiveness; etc. In this concern, coherence may be viewed as minimizing the antinomies and normative gaps and discarding untidiness produced by politics or inconsistencies provoked by sound moral doctrines⁹¹ .

History has taught us that the appeal to dogmas as criteria of truth only leads to disastrous outcomes, irrespectively of whether the conformity occurs in the name of religious beliefs or political ideologies. In law, and also more generally, the appeal to nature has been a flexible formula: it has been claimed as propaganda by some philosophers to ground the possibility of fighting the regime or support it. In attacking axiomatic systems of natural law and the arbitrary character of metaphysical assumptions, Ross claimed that «like a harlot, natural law is at the disposal of everyone »⁹² . The unsatisfactory outcome of displaying unified legal systems has been keenly maintained by Eckhoff who rests his claim on the fact that «it is psychologically understandable that many people have sought for such an anchorage in common for all legal norms. It satisfies needs for simplicity and for considering the law as unity. And one gets a facile way of justifying the validity of legal norms. But the justifications that are provided by operating with a highest norm common to the legal system are more facile than convincing »⁹³ .

A possible alternative to axiomatic systems hinged on deductivism is defeasible reasoning. While in a conclusive reasoning one can always adopt a conclusion y while endorsing x premises (y should never be rejected while endorsing x), in a defeasible reasoning «one should, under certain conditions, refrain from adopting its conclusions though endorsing its premises »⁹⁴ . From the truth-perspective, whilst conclusive reasoning are truth-preserving in that if x are true, y is necessarily true, thus providing a monotonic logic, on the contrary defeasible reasoning are not truth-preserving in that y may be, under certain circumstances⁹⁵ false even if x are true. As a peculiarity in human reasoning, defeasibility is a notion invoked in various contexts, from epistemology, to moral philosophy, to artificial intelligence, to legal theories. By following in the footsteps of Pollock and Sartor, I have already illustrated examples of defeasible inferences (see: section 4). The defeasibility perspective is endowed with the idea of limited factual and normative knowledge: this does not in any way run counter to the fact that the agent must be able to formulate reasonable decisions though the unlimited amount of information which he could not be aware of. Accepting defeasibility in the law and legal theory best suits to capturing important features both in the knowledge of law and in adjudication. «The legal reasoner – as Prakken and Sartor say –though endowed with limited knowledge, energy, time and resources (…) a defeasible legal reasoning (…) should enable a legal agent to form judgments on the basis of the knowledge he has, and the thinking he is presently able to do, and correct (and possibly withdraw) such a conclusions as soon as he is able to take into account further legally relevant information »⁹⁶ . If on the one hand defeasibility contrasts with any attempt to reduce representation of the law to consistency, on the other hand defeasibility does not entail a chaotic picture of legal knowledge and the law. As to the former, though each legal theory includes or should include inconsistent and conflicting knowledge, this «should nevertheless provide an adequate framework for legal thinking and problem-solving »⁹⁷ . Legal theories are defeasible but from this it does not follow that new knowledge retracts previous knowledge, in virtue of the dynamic adjustments and the capacity to address new issues. With reference to the latter, although qualitative dimensions of value may generate incomparability and make it fruitless to engage in comparison and weighing, one must acknowledge that law incorporates moral precepts and developscoherent bodies of moral doctrines. The fundamental interplay between moral values and criminal and procedure law, the cross between moral agency and legal responsibility, the core of the moral doctrines inspiring Constitutions and the body of human rights or, again the establishment of moral precepts through adjudication of law, acquire such a strong relevance in the understanding of law, in legislation and application of law. Against the spirit of philosophers who have yielded to a perennial temptation to postulate a realm of ultimately deductive truths, the definition of law is a flexible concept itself depending on an intrigued pattern of moral and legal reasons. For purposes of knowledge, we must observe that law requires immediately justified propositions whose epistemic status is positive: the authority argument and the consequent hierarchy of sources. Sousa e Brito deploys an argument that legal sources (such as Constitution, statues, jurisprudence) display logical priority in legal argumentation and this entails the setting of hierarchical relationships among the sources of law. The authority gained by the legal sources generates an hermeneutical priority among the sentences of the sources in the sense that the legal sources «may have been used merely as starting points to discover a logical system by which the decision is justified »⁹⁸ . This provides a basis for seeing how a legal system does not only consist in deductively arguments inferentially justified and whose premises are incorrigible, thus it allows for a concept of justification which is double-aspect and essentially gradational. A relevant result of this perspective is that the hermeneutical priority of the sources eases the thickness of social fact thesis (according to which only social facts generate law) and accommodates Rawls’ concept of reflective equilibrium as the space where «conviction of any level of generality may provide supporting reasons »⁹⁹ . Brito’s concept of hermeneutical priority may accommodate the basic argument of legal defeasibility as the one of Prakken’ and Sartor’s: «legal rules and principles are to be viewed as reasoning warrants that, when certain antecedent reasons are available (…) support the derivation of legal conclusions. In this way basic legal inferences are obtained which consist of chains of warrants leading to legal conclusions »¹⁰⁰ . Furthermore, Brito’s supporting reasons-argument is comparable with the argument-based model of defeasibility resulting from the interaction of conflicting arguments. As appears above, according to Pollock’s epistemological model of defeasibility any argument may be defeated, rebutted or undercut by further arguments; in turn, the winning argument may be contradicted (and defeated) by further arguments and this may generate a recover of the previously defeated argument. It does follow that this model of legal reasoning grounds on – in Sartor’s words – «the dialectical interaction of competing inferences: the outcome of this competition determines what conclusions will be legally justified in the framework of available legal knowledge »¹⁰¹ .

The dialectical interaction between arguments is especially useful to appropriately account for adjudication of law, and more particularly the legal dispute where a proponent interacts with an opponent. The relational dialectic in adjudication gives rise to a dialectical play of patterns and roles which embodies the defeasible model of legal argumentation. In the light of the currently available knowledge, the plaintiff exhibits an initial argument which is criticized by the critical defendant. The argument may be defeated in case of the defendant uses an argument to which the proponent is incapable to reply; or it may be undercut by further arguments of the defendant; and in turn, it may be contradicted (and defeated) by further arguments of the plaintiff in such a way that the initial argument may be reinstated. The dispute takes the form of a game which ends when one of the reasoners successfully finds an «argument game» against which the other player is unable to counterargument¹⁰² . As said earlier, this model is far from an exclusive application of logic rules: for a reconstruction of a legal dispute, we take ourselves to know that legal arguments must also be modeled on the basis of procedural limits which render the legal reasoning not purely logic. If one considers the arguments which are generally used as correct in legal discussion, we will notice that they show limits which render the practice of law not purely logic. A great deal of legal arguments fall under a «procedural filter» which one assumes is of significance for the conclusion of the dispute¹⁰³ . The law of decision making procedure regulates the burden of proof, the admissibility of evidence or counterevidence, the number of sentences admitted in legal argumentation for reasons of practicability, time-saving, security and above all democracy, for instance¹⁰⁴ . Concerning the process-based defeasibility, Prakken and Sartor say that «this process evolves in time: the arguments that pass the filter at a certain stage in a dispute may favour the plaintiff, but the dispute may proceed (e.g. the burden of proof may switch to the defendant) and a new set of arguments passing the procedural filter at a large stage may favour instead the defendant »¹⁰⁵ . In the given quotation lies a good deal of material for illuminating the aspects of the defeasibility of legal reasoning and the narrowness of the inference-based account of argumentation in adjudication of law.

The pivotal idea of the work is that – in respect of the great diversity and vagueness in the way the terms «foundationalism» and «coherentism» are used in various contexts and the uncomfortable usage of drawn dichotomies – both theories of justification can only be thought to be rivals. Actually, they both join the top-down model of deductivism. As holism can relieve people from contradictory impulses and gratify their desire in global coherence, so happens with legal systems, and more generally with any kind of intellectual system. But this reveals not more than an attempt to switch idealistic views of law. In systematizing law, jurists’ tireless work seldom engages in global interpretation of the whole system and, more importantly jurists do not purport to systematically resolve any given conflict since no best answer is admissible. The same goes for law adjudication: the judge is not a volitionalmonstrum who chooses the best resolution between ideas of justice and fairness. Although the pursuit of coherence is irresistible, the doubts with regard to the capacity for coherence to capture the irreducible complexity of our lives as well as the tangled architecture of legal systems are inescapable.

I have sought to demonstrate this through three related avenues of investigations. First, by letting the reader see the applicability of foundationalism and coherentism in epistemological domain. Secondly, by letting the reader see the applicability of coherentism in the legal domain by taking Dworkin’s metaphoric figure of Hercules judge into special account. Thirdly, I have adopted the very general idea of defeasibility employed in contemporary theories of knowledge in order to highlight the dynamism of legal reasoning. In this concern, defeasibility requires degrees of justification and degrees of justification work like probabilities: «a correct account of defeasible reasoning – Pollock said –requires us to appeal more seriously to degrees of justification »¹⁰⁶ . Defeasibility in law allows the relevance of limited human capability and, with reference to the unsurmountable and constantly changing mass of legally relevant arguments, acknowledges a double aspect conception of justification, i.e. the inference-based account and the gradational peculiarity of justification.

Defeasibility is, however, problematic. This is so especially because it is difficult to find any tenable grounding for such «dialogical games», given the indeterminacy in giving criteria for forms and degrees of justification. In recent legal philosophical literature, a conspicuous number of authors have traced the same indeterminacy in providing fixed criteria for what counts as a good legal argument, and what strategies are indispensable for implementing a good legal argument¹⁰⁷ . A large part of the question of what criteria are relevant in the assessment of defeasibility can only be answered by going to a higher level of abstraction, which consists in comparing various theories, since no one right theory is possible to get an overview of all relevant factors in justification. The more difficult one considers it to be to get an overview of all relevant factors, the more one is forced into a particular form of specifying criteria that build its own theory. The more theories of the same legal domain are possible, the larger is the reason to warn that ways of comparison and selecting the most appropriate theory is a crucial need¹⁰⁸ . In actually occurring investigations, coherence is again the most popular and evoking word which makes it easier for one to catch sight of the fact of appropriateness of a body of knowledge. A commonly held opinion is that coherence involves «comprehensiveness»: Peczenik held the opinion that «the more the statements belonging to a given theory approximate a perfect supportive structure, the more coherent the theory is »¹⁰⁹ . The same view is central in Alexy: what renders a theory tenable is the sufficiently «big size» in the sense that «a coherent set of propositions should comprise as many as different propositions as possible »¹¹⁰ . Nevertheless, Alexy himself has pursued a suspect toward coherence as a genuine superior criterion, and by this he meant that coherence itself demands a recourse to additional criteria. To avoid the problems posed by coherence and in order for a system to be complete, Alexy admitted that coherence should be combined with the procedure of rational discourse and history¹¹¹ .

The insight to be gleaned by my observations is that since theories are defeasible for the same reasons which can affect the semantics of human reasoning, one should endorse that «a coherent legal theory should be more than an internally connected web of concepts »¹¹² , thus increase the attention to howto measure the coherence among competing theories. This gives us a reason for thinking coherence as an ideal encouraging the investigation of legal scientists (theorists and judges), more than a super criterion for an adequate choice of the best theoretical alternative.

---

Prof. Dr. Alessandro Serpe, University «Leonardo da Vinci» of Chieti

a.serpe@unidav.it

Alessandro Serpe (born 1977) received a PH.D. in legal philosophy (Naples Federico II University/2002) and carried out his researches at the University of Oslo under the guide of Prof. Svein Eng. From 2007 until 2011 he worked as Post-Doctoral researcher in legal philosophy (Lisbon New University) under the supervision of Prof. José de Sousa e Brito. Since 2001, Dr. Serpe had been constantly lecturing on legal philosophy at the Federico II University of Naples. From 2004 he is Guest Professor at Zulia University (Venezuela) and from 2011 he is Visiting Professor at Carl von-Linde Akademie, Technische Universität München (Germany). Currently, he is research fellow at the University «Leonardo da Vinci» of Chieti.

Dr Serpe’s scientific interests concern legal philosophy and general legal theory with special reference to analytic legal philosophy, theory of argumentation and normative ethics.

My thinking about these issues has been influenced by wonderful discussions with wonderful legal philosophers. I owe a special debt of gratitude to my Mentor, Professor José de Sousa e Brito, who has been a constant source of unerring support and philosophical insight over the last years spent in Lisbon as Post-doctoral researcher. I must admit to have been keenly affected over the years by his intellect and voluminous knowledge. Moreover, I can truthfully express my appreciation to Professor Giovanni Sartor for having helped me in clarifying my thoughts and provided me with important literature. The international Conferences QAJF 2010 and QAJF 2011 have been excellent places of valuable suggestions in examining these issues.

---