What is the normativity of logic? If logic is understood as “a specification of a relation of logical consequence on a set of truth-bearers,” where consequence relations “preserve truth in virtue of logical form” (Steinberger Reference Steinberger and Zalta2017, 2), it is natural to think that logic has normative significance. For instance, noting that we often make errors in reasoning, one might assume that logic is not about how we do think but about how we ought to think. However, such a view runs into the problem that logical laws seem to say nothing about thinking; they are purely formal, while thinking is something we do, whether occasionally or all the time. How does a logical law gain normative traction on apparently psychological acts such as belief-formation or reasoning? One might think that it does so by means of a hypothetical imperative: If you want to preserve truth in such acts, then you ought to reason in accord with logical laws. Logic would be normative in the sense of providing evaluative norms that represent success-conditions for a practical project, that of preserving truth. But this leaves another normative question open: Why ought we to preserve truth in our thinking? And while the answer to this question may seem obvious, it can come into conflict with other desires we have, and so with other norms that pertain to them, such that, in some cases, thinking or reasoning or believing or asserting ought to ignore ((p ⊃ q) & p) ⊃ q.
To take one example, Gilbert Harman (Reference Harman1984) argued that logical principles are not directly normative for a theory of reasoning or belief-revision. To believe that p and that p ⊃ q does not mean that one should believe q. Perhaps the preponderance of evidence speaks against q.Footnote 1 One response is to hold that logic supplies the constitutive norms for reasoning: Just as, in chess, if one consistently tries to move a knight diagonally, one does not count as playing chess, so too, if one continually embraces invalid inferences, one does not count as reasoning at all. Of course, I can make mistaken moves, but as long as “I acknowledge that my activity is answerable to the rules,” then I still count as playing (Steinberger Reference Steinberger and Zalta2017, 14). Steinberger continues: “It is not easy to specify … what the requisite acknowledgement or sensitivity consists in”; but it is just this, I hope to show, that is provided by Heidegger’s phenomenological ontology of reasoning.
The interesting thing about constitutive norms is that they do not involve any “ought,” so logical laws would simply specify what reasoning is. As Emanuela Carta (Reference Carta2021, 130) puts it, “contrary to deontic norms, constitutive norms do not engender obligations or place demands on agents to act or behave in certain ways.” But with this we are back to the beginning: Reasoning just seems to be a game that we can either play or not play. The “normativity of logic,” however, “does not seem to be optional in the same way” (Steinberger Reference Steinberger and Zalta2017, 8). And were one to argue that reasoning is somehow essential to us – that we cannot “be” without playing the game – then it needs to be explained how we can still be what we are even when we fail to reason logically. Beyond the game analogy, this is an ontological question that brings us to the threshold of Heidegger’s approach to the normativity of logic.
I will argue that Heidegger, like Kant, provides an account of the constitutive normativity of logic; however, what logic is constitutive of is not something that is essential to us in the way it is for Kant.Footnote 2 Rather, it is constitutive of a practice that, for Heidegger (at least in the period covered in this chapter, up to 1930 or so), stands in a certain opposition to his ontological project, whose goal is not to explain entities and their properties but to clarify the “meaning of being” phenomenologically. Because phenomenology initially stands on an equal footing with logic in the project of ontology, it has normative parity, and so it is possible, as Heidegger famously claimed, that the “reign of logic … disintegrates in the turbulence of a more original questioning” (Reference McNeilGA9, 117/92). Of course, this introduces its own asymmetry between logic and phenomenology, a point to which we will return.
So, the “reign” or normativity of logic may not be absolute. Nevertheless, I will argue that, even for Heidegger, logic is not exactly optional either. Logical laws are constitutive norms of reasoning, and though we are not, “essentially,” rational animals, reasoning is something we not only can do but are obligated to do thanks to what Heidegger calls the care-structure of our being. The ontology of care provides a normativity-first account of reason that shows why logic is more than a game: Logic formalizes “what we owe to each other” (Scanlon Reference Scanlon1998) and belongs to the practice of coming to mutual understanding (Habermas Reference Habermas1990).
1.1 Contemporary Discussion of the Normativity of Logic
Before turning to Heidegger, it will be helpful to look briefly at how discussions of the normativity of logic that do not proceed phenomenologically understand the motivations for thinking that logic is normative, and how they frame the objections that have been raised to this idea. Gillian Russell examines three such motivations – the argument from “normative consequences” (NC), the argument from “error” (E), and the argument from “demarcation” (D) – on the way to arguing that logic is a purely “descriptive” enterprise that is only “weakly entangled” with normativity, that is, logic is normative only when combined with “widespread background norms about the relations between belief, reasoning, and truth” (Russell Reference Russell2020, 387).
(NC) hold, roughly, that “is valid” or “is a logical consequence of,” when used to describe an argument-structure, is epistemically normative: It tells us what we ought to believe (Russell Reference Russell2020, 374). (E) appeals to the difference between descriptive laws, which are never violated, and norms, which are often violated, to argue that since we often fail to reason logically, logical laws cannot be descriptive. They must be normative. Finally, (D) addresses the worry that if logic is neither psychology nor semantics – neither the study of how we actually think or reason nor the study of truth-preservation in natural languages – then there is nothing for logic to be about. But if logic is about how we ought to think or reason, then it has a distinct domain (378).
(D) also addresses a worry that logic is merely conventional. If a classical logician accepts ¬ ¬p ⊃ p and an intuitionist logician rejects it, and if both systems can offer model-theoretic accounts that support their views, then either we can say that they “mean different things” by the entailment relation or we can ask which of them is the “real” entailment relation. But on a descriptivist approach, there might be no answer to this latter question. For instance, Steinberger (Reference Steinberger and Zalta2017, 4) notes how this descriptivist result is reflected in Hartry Field’s view that “validity” is primitive and that its normative meaning is determined by the “conceptual role” it plays within competing logical systems as a constraint on our “doxastic attitudes.” On a normativist account, in contrast, we can ask, of either logic, whether it is how we ought to reason.
(D) involves what Russell calls the “strongest” degree of logic’s “entanglement” with the normative. Normativity is (at least partly) constitutive of what counts as a logical theory. Disambiguations of “validity” (as in the dispute between classical and intuitionist logicians) do not count as logical theories unless they address what is demanded in (D): Which one is normative for how we ought to reason? (Russell Reference Russell2020, 379). Constitutive norms, as we saw, do not involve an “ought” in their formulation; nevertheless, their function cannot be understood without recognizing their normativity in regard to how something is supposed to be done. That is to say, constitutive norms yield evaluative norms for success in the “game” we are playing. While both “deontic” norms and “evaluative” norms include an “ought” in their formulation, only deontic norms constitute obligations on a specific addressee, while evaluative norms derive from the eidos of the type in question (Carta Reference Carta2021). For this reason, evaluative norms are best expressed as “should be” (Sein-sollen) rather than as “ought to do” (Tun-sollen). Thus, the constitutive norms of a practice entail evaluative norms for assessing success in living up to the eidos of being a player; they do not demand that a player play a certain way.
A weaker view, represented by (NC), is that logic is entangled with normativity because it has direct normative consequences for belief-formation and revision (for instance, if one takes “correctness” to be a normative notion). Weakest of all is the idea that logic has normative consequences not on its own but only “alongside other (perhaps quite prevalent) normative assumptions.” If you make an error in arithmetic, for instance, then your sum is false. But this has normative significance only “in conjunction with the non-arithmetical normative fact that one ought not to believe false things” (Russell Reference Russell2020, 380). Because the prior normative commitment does all the work, logic remains a descriptive theory. Like physics for mass and heat, logic simply tells us what “is” in regard to entailment or validity. The “widespread intuition that logic is normative” is explained by the fact that “when we make claims about entailment we are commonly also intentionally conveying normative information as well,” and this conveyance “has a tendency to stick to logic” itself (382).
So, what does such a descriptivist position look like, and what answers does it provide to the three motivations for thinking that logic is normative? Russell’s account will position us to see how a phenomenological approach – Heidegger’s in particular – can put the debate on a new footing.
For Russell, “logic is the study of patterns of truth-preservation [e.g., entailment] on truth-bearers” (382). Thus, like physics, it has an object-domain (“sentences”), and just as physics studies certain properties of objects, logic studies how “the property of truth” is preserved in argument-forms (382). Taking logic to be about sentences can most easily be seen in artificial or formal languages, but Russell assumes that natural languages have such descriptive properties as well, though perhaps difficult to capture in formal notation. This is only an “epistemic difficulty,” however, no more worrisome than are similar difficulties in applied physics (385n15).
On such a view, the arguments for the normativity of logic are easily dispensed with. (NC) is undermined because to say something is “valid” or “entailed by” is simply to say something about truth; any normative consequences come from a “conjunction with norms linking truth to belief and reasoning” (385). (E) is undermined by the observation that “there are no violations of logical laws” and cannot be any, since they are “descriptive laws of truth.” Supposed violations of such laws are simply mistakes: I misapply or misunderstand the law, but I in no way “violate” it, since the law has nothing directly to do with application (386). Finally, (D) fails because, while it may be difficult to determine “whether or not an argument form is valid” for a given language, answering this question does not require us to consider whether it is “normative” for reasoning. We just need “better semantics,” and ultimately the right model theory for it – that is, “more theory and evidence,” not a normative conception of logic (386).
With this descriptivist account of logic in mind, we turn now to Heidegger, whose phenomenological approach to the connection between logic and truth gives us reason to accept the main claims of descriptivism but also to challenge the idea that appeal to “widespread background norms about the relations between belief, reasoning, and truth” (387) can adequately account for the kind of normativity that is intrinsic to logic.
1.2 Logic Is Not a Normative Discipline
From his earliest days as a student, Heidegger had an interest in logic. Though aware of the “logistics” of Russell and Whitehead (GA1, 41–42), Heidegger’s attention lay elsewhere – namely, in the demarcation argument (D). In what domain do logical laws hold? If the laws of logic are supposed to pertain to thinking or reasoning, then they might appear to belong to psychology, suggesting that they are, at least in part, empirical and contingent. In his dissertation, Heidegger draws on Husserl to show that various current psychologistic theories of judgment misunderstand the character of logical laws, their a priori “validity” or “holding” (Geltung). Similarly, he argues that logical laws cannot be grounded in the “grammar” (vocabulary, sentence structure) in which judgments are expressed, since different languages express the same thing with different words and grammatical structures. Heidegger’s initial attempt to address (D) thus settles on the idea that the domain of logic is the judgment’s ideal meaning, the “proposition.” Validity is a property of meaning (Sinn), and logic thematizes the entailment relations among propositions, given their categorial forms. The relation to truth is taken to lie in the fact that such judgment-meaning “holds” of the things about which the judgment judges. Logic is not a normative discipline but a theoretical one.
The details of this approach to the domain problem – which then as now are widely accepted – need not concern us, but a glance at Husserl’s argument for it in the Logical Investigations (Husserl Reference Husserl1970, 74–89), which Heidegger adopts, will make the latter’s eventual departure from it more perspicuous. Husserl’s general argument is that all normative disciplines presuppose theoretical disciplines. For instance, to say that “a soldier ought to be brave” presupposes that we know what descriptive properties a soldier has, those properties that make up what it means to be a soldier (essence, ideal meaning). If bravery is among them, then we can always transform the descriptive statement “A soldier is brave” into a normative statement, “A soldier ought to be brave,” where the “ought” here is an impersonal evaluative “should” expressive of the success-conditions for a certain type (here, soldiers). But this does not mean that our cognition of what a soldier is is intrinsically normative or presupposes a normative discipline. In the case of logic, the claim that in accepting p and p ⊃ q one ought to accept q does not stand on its own; it is justified by the purely descriptive relations that hold (are valid) between the ideal meanings of p and q, that is, their status as true, and in that regard (as Harman reminded us), evidence can trump reasoning. Logic is not a normative discipline.
However, while appealing to Husserl’s position throughout his dissertation, Heidegger concludes by turning the argument in a different direction: “What is the meaning of meaning?” “What is meaning?” (GA1, 170–171). By 1923, Heidegger had come to question Husserl’s starting point on phenomenological grounds: By starting with “already known knowledge,” that is, with ideal meanings (propositions) taken as elements of logically structured sciences, “truth” is taken to name a property of propositions, that is, their “holding” (Geltung) of their objects (GA17, 58–60; Dahlstrom Reference Dahlstrom2001, 131–138). If ideal meaning is truth-functional in this sense, and if there are formal laws that preserve the property of truth through various transformations of meaning, then logic is the theoretical discipline that identifies such truth-preserving entailments. But Heidegger came to think that this schema is phenomenologically superficial: While it may appear, when one begins with developed sciences, that an ideal meaning’s holding of an object is one of its properties, it is not. And if that is so, then the question of logic’s normativity cannot be decided by appeal to Husserl’s distinction between normative and theoretical disciplines, that is, by starting with already known knowledge. It demands a phenomenology of truth and so of “truth-bearers,” one that illuminates what is “preserved” in logic.
1.3 The Phenomenological Ground of the Truth-Predicate
If the aim of Heidegger’s early work can be described as an attempt to liberate logic from grammar – that is, to identify the logical elements of ideal meaning behind the surface grammar of sentences – Being and Time calls for “liberating grammar from logic” (Reference HeideggerSZ 165), an account of meaning that does not begin with the phenomenology of cognition. The supposed ideality of meaning – “validity” as the “object” of logic – is now declared to be “very questionable” due to its opaque because equivocal character (Reference HeideggerSZ 156): validity is understood ontologically as the “ideality” of judgment-meaning; epistemologically as the “objectivity” of that meaning, its “holding” of an object; and pragmatically as the “universal” validity, the “bindingness” of objective meaning “for everyone who judges rationally” (Reference HeideggerSZ 156).
Heidegger does not dismiss any of the phenomena that this equivocal term references, but he places “no advance restriction” on the “concept of ‘meaning’” that would tie it to this whole “logical” problematic. At bottom, his target is the “unclarified separation of the real and the ideal,” and he argues that “psychologism [is] correct in holding out against this separation” just because it is unclarified (Reference HeideggerSZ 217). However, Heidegger’s point is not that logical psychologism is correct; rather, it is that there is a need for a phenomenological account of meaning, judging, and reasoning that does a better job of elucidating the ontological ground of cognition or “already known knowledge.” And in pursuing this task, Heidegger suggests a different account of the normativity of logic: Logic is constitutive for a certain practice (hence it is descriptive, involving no “ought”), but the practice itself, reasoning, is normatively demanded of the kind of being we are.
The first step in Heidegger’s phenomenological account is found in The Basic Problems of Phenomenology. In his criticism of Hermann Lotze, whose Logik introduced “validity” as a Zauberwort, we can clearly see why, for Heidegger, truth is not, ontologically, a descriptive property of sentences.
To treat sentences as truth-bearers is to construe truth as a property of sentences. The sentence “Today, the sky is blue” has the property of being-true such that “The sentence ‘Today, the sky is blue’ is true” has the same logical form as the embedded sentence. As Heidegger shows, Lotze’s account of the ideal meaning of such sentences rests on the centuries-old tradition, deriving from Aristotle, which holds that the “is” (or “copula”) represents a kind of “combination” of subject-term and predicate-term. For this reason, Lotze maintained that “a negative copula is impossible,” since nothing is combined in it, and this suggested to him that all judgments are really “double judgments.” Thus, “S is p” should be understood as saying that “‘S is p’ is true,” while “S is not p” says that “‘S is p’ is not true” – making explicit, in each case, that truth is or is not a property of the sentence.
This solution, which itself involves a sentence with a negative copula, threatens an infinite regress of such property-attribution, but if we respond by treating negation as a logical constant, moving it “outside” the sentence, ¬ (S is p), or by replacing the “combinatory” copula with the existential quantifier, ¬ (Ǝx)(Px), we conceal what Heidegger sees as the primary ontological problem raised by logic: What mode of being (Sein) is represented by the copula?
Lotze answers that the copula represents the judgment-meaning’s holding (being valid) of its object as a known object – its “objectivity” or “being-true” – and that this is co-intended in the copula by means of a “subsidiary thought” (GA24, 287–288). Heidegger, in turn, accepts the idea that the copula expresses the being-true of the assertion and that in “every uttered assertion [sentence] the being-true of the assertion is itself co-intended,” but he denies that this co-intention is a “subsidiary thought” that attributes a property to something: “Truth is not a being that appears among other extant things” (GA24, 304). It is “never extant like a thing but exists” (GA24, 310).
So, in what way is truth co-intended in the assertion? Heidegger’s answer goes back to Aristotle’s claim that truth is not in things but en dianoia, in the “understanding.” This does not mean that truth is a property of (some) thinking; it means that what truth is can be approached only “in the middle ‘between’ things and Dasein” (GA24, 305). For our purposes, the result is that truth-bearers cannot be understood as any kind of entity. Heidegger hints as much in Being and Time: “Assertion is not the primary ‘locus’ of truth” (Reference HeideggerSZ 226). Put positively, if we want to understand how logic can be the descriptive science of formal sentential truth-preservation, we must first understand how assertions, and so sentences, can be truth-bearers at all. For Heidegger, this is an ontological problem and so requires phenomenological reflection on the “being” of assertion. And central to his approach is the claim that assertion cannot be understood without taking its “reference to things” into account (GA24, 280).
This reference to things cannot be understood if we start with the “verbal sequence” (sentence) in which an assertion is expressed; rather, we must phenomenologically describe “the specific contextual interconnection” that belongs to the phenomenon: the “relational whole of word, signification, thinking, what is thought.” And within this interconnection, the key point is that asserting is something we do; it is “one of Dasein’s intentional comportments” (GA24, 294–295). A comportment is that for the sake of which (Worumwillen) we do what we are (currently) doing, so in asserting something, we act for the sake of being a truth-teller: “Prior to the assertion and for the sake of making it, the asserter already comports himself toward the relevant entity and understands it in its being” (GA24, 300, emphasis added). For this reason, Heidegger concludes that “being-true already lies in assertive comportment itself.” It is co-intended, but not as a property of the assertion; rather, “truth exists. Truth possesses the mode of being of Dasein” (GA24, 313). But with this, we are still far from understanding the relation between logic and assertion, so we need to move a bit more slowly.
Heidegger’s phenomenological analysis of assertion highlights three moments: “assertive comportment” is (1) apophansis, “exhibition”: “letting something be seen as it is in itself.” Apophantic exhibition is (2) predication: a “sundering” (Auseinanderlegen; dihairesis) of “the belonging-together of the manifold determinations of the being which is asserted about” so as to characterize it in a particular way (synthesis). Finally, assertion is (3) communication: an “understanding comportment toward the being about which the assertion is being made” in which that comportment (and not mere information) is or can be shared with others. In phenomenological terms, then, assertion is “communicatively determinant exhibition” (GA24, 297–300).
For Heidegger, each of these characteristics points to the fact that assertion is not our original access to things but is “related to something antecedently given as unveiled” (GA24, 296), that is, to something available as understood. Assertion exhibits something already there and, by sundering the manifold determinations that “belong together” in it so as to give it a definite determination, exhibits it in such a way that the comportment of truth-telling can be explicitly shared. Thus “assertion does not have a primary cognitive function but only a secondary one. Some being must already be unveiled” – available as understood – “if an assertion about it is to be possible” (GA24, 299).
If the “is” of assertion is understood or co-intended as “being-true,” then truth is not a property of the sentence taken as a truth-bearer; rather, it “signifies a being in its unveiledness” (GA24, 303), its availability for assertive comportment. This being-available for communicatively determinant exhibition is what “bears” truth: Truth is en dianoia, in the understanding, not as a property of that understanding but as the availability of an entity in its organized manifold (meaning); and truth is also “in things,” not as a property but as “a determination of the being of the extant, extantness” (GA24, 305). Truth, unveiledness – “‘between’ things and Dasein” (GA24, 305) – befalls things thanks to Dasein’s “understanding of being.” Here, we can begin to see why phenomenological reflection on intentional correlation, the “original” bearer of truth, has a certain priority over logic in an ontological investigation of what truth is.
The phenomenological picture, then, is this: The relation of assertion to truth does not involve positing a descriptive “third realm” of “valid meaning” (Husserl, Frege, Lotze) – truths in themselves – but is grounded in the ontology of Dasein as comportment. An assertion is true if its sundering-synthetic determination exhibits the entity as it is in itself. But such exhibition depends on the prior meaningful availability of the entity: Assertion “is a mode in which Dasein appropriates for itself the uncovered being as uncovered” (GA24, 312). Acting for the sake of being a truth-teller is appropriation, making one’s own, but such comportment is possible only because Dasein itself is “unveiled to its own self for itself” (GA24, 308), that is, because Dasein “understands” both the being of things and its own being – in this case, what it means to be a truth-teller. As Heidegger puts it: “We take being-true in this wholly formal sense as unveiling” – that is, as aletheuein, “to take out of concealment” – with the result that Dasein “is true.” But “being unveiled” is not a property of Dasein; it is its “way of being” (GA24, 307–308). In being unveiled to itself in its comportments, Dasein uncovers available entities such that they can be “encountered within the world,” that is, “disclosed for the existent Dasein” in those comportments (GA24, 313).
In Section 1.4, we will examine how, exactly, Dasein is disclosed to itself in its comportments, and this should clarify the ontological ground of logic’s normativity. But already we may note one consequence of Heidegger’s approach. Heidegger argues that we may call an extant entity “true” not “intrinsically” but “as uncovered in the assertion” (GA24, 312). Now, even though language is not an extant entity but “is as Dasein is,” that is, “exists” (GA24, 296), a sentence is an extant entity – a structured string of written or spoken words – and so it is not intrinsically a truth-bearer but only “as uncovered in [an] assertion” about the sentence. Just this happens when logical theory asserts that truth is a property of sentences, but on Heidegger’s view such an assertion is false. As Heidegger puts it, if assertions yield sentences, not all “sentences” can be “traced back to theoretical statements without essentially perverting their meaning” (Reference HeideggerSZ 158). Nevertheless, even if the sentence attributing the property of truth to sentences is false, logic might still be understood as describing truth-preserving entailments among sentences. In that case, though, what is “preserved” will not be a descriptive (“real”) property of sentences. And if that is so, we have not yet decided the question of the normativity of logic when we have described such entailment relations. We would still need to reflect on the ontological ground of assertions and the sentences that express them, a ground that might yield a different understanding of the normativity of logic.
1.4 The Normativity of Meaning: “Is” and “As”
Being-true is co-intended in the assertion, not as a double judgment but as constitutive of assertive comportment. But the assertion itself can be “either true or false.” What accounts for this possibility? In Fundamental Concepts of Metaphysics, Heidegger provides an extensive answer to this question, from which we will here draw only a few points relevant to the question of the normativity of logic.
In Being and Time, Heidegger argued that meaning does not originate in asserting but in what he calls the “hermeneutic ‘as’ of interpretation,” the way we encounter something meaningfully as something: as a hammer, for instance, or as too heavy (Reference HeideggerSZ 158). Such interpretation “articulates” – that is, “takes apart” what it also “binds together” – that which is understood in my comportment. In saying something about what is so encountered, the hermeneutic “as” is reduced to the “apophantic as”: The hammer is heavy. The “specialty of the assertion” (Reference HeideggerSZ 158) is thus to exhibit things in the “dimension” (GA29/30, 425) of the extant, the presence of a property “in” an object. This reduction is “formalized” as a “relation” and the assertion is thereby “dissolved logistically” and “becomes the object of a ‘calculus.’” However, this conceals the way that logic is “rooted in the existential analytic of Dasein” (Reference HeideggerSZ 160). Only if this existential rootedness is excavated phenomenologically can we understand the sense in which logic is – and is not – normative.
The goal of Heidegger’s phenomenology of logic is to show that the “is and the as have a common origin” – namely, Dasein’s “fundamental comportment: being free in an originary sense” (GA29/30, 497). Unpacking this claim will show how the logical calculus is a descriptive theory of truth-preserving entailment relations, not a normative theory of how we ought to think. But it will also show that logic is not normatively optional for us either, since its laws are constitutive norms of a practice, reason-giving, that is demanded by our mode of being: “being free in an originary sense.”
Heidegger’s problem with formalization is that it eliminates the “dimension” in which a being is encountered “as” it is (GA29/30, 435–440). I encounter the hammer as too heavy when I find it unsuitable for the task at hand – that is, within the dimension of the available, the meaningful context (Bedeutungsganzes) in which the hammer shows up in its “equipmental” being.Footnote 3 The assertion abstracts the hammer from this dimension and constitutes it as an object possessing the property of heaviness. There is nothing erroneous about such abstraction, but it gives rise to the illusion that the subject-predicate form is sui generis and that the “meaning” or proposition expressed in the assertion can, by itself, be the bearer of truth. But, for Heidegger, such meaning is parasitical: “A is b” only has meaning thanks to the experience of A as b (GA29/30, 436).
Now, this sort of experience – the hammer as too heavy – is itself possible only if the hammer is encountered in a normatively structured context of other things in which it fails to serve as it ought to serve. Such contexts are possible only if I “hold myself in a comportment” in which I am “able to refer to [meinen, intend] these other beings as such” (GA29/30, 448). This same point holds of assertion as well: Since asserting is “an essential manner of comportment” belonging to Dasein (GA29/30, 486), it too is normatively suspended between success and failure. The “essence” of assertion, for Heidegger, is its possibility of being “true or false,” such that no assertion can simply be “true” all by itself (GA29/30, 449). And if that is so, then the “falsity” of an assertion, like the “too heavy” of the hammer, depends on the context disclosed as meaningful in the kind of comportment that allows assertive reference to other beings – namely, the comportment of being a truth-teller. “Truth,” then, is not a property of an assertion but a success-term, an evaluative norm that derives from the constitutive norms of the comportment of truth-telling. To see this, we need to take a closer look at Heidegger’s phenomenology of comportment.
Comportment (Verhalten) is acting “for the sake of” (hou heneka) some possibility of being; that is, it is trying to be something: teacher, father, citizen, nurse, scientist, and so on.Footnote 4 To “try” in this sense is to care about succeeding or failing at one’s practical identity – that is, to act in light of the norms that govern such success or failure. These norms are both evaluative (measuring the extent to which I succeed) and constitutive (measuring whether I count as trying-to-be at all). If I don’t have the ability (Seinkönnen, “possibility” in Heidegger’s sense) to do what is demanded of teachers or carpenters, then no matter what I do, I don’t count as acting for the sake of being a teacher or carpenter. “Comportment,” then, formally indicates the being of Dasein as care (Sorge): that being “for which, in its being, that very being is essentially an issue” (Reference HeideggerSZ 84). Put otherwise, Dasein is only in trying to be something.
The norms that govern practical identities are, in the first instance, public: One acts as one does around here. I “understand” what it means to be a teacher because teaching is an existing practice where I am from, a practice that, in complicated ways, reflects a common understanding of what teachers are supposed to do. But this normative “supposed to do” does not provide any recipe for being a teacher; it is not a technical or instrumental norm that can be expressed as an in-order-to. My success in building a birdhouse is measured by the result (the “work,” ergon), and the work determines (in principle) the steps I should take to get there. In comportment (hou heneka), in contrast, success or failure is measured at every moment – which is another way of saying that the meaning of teaching, what a teacher should be, is always at issue.
Public norms institute criteria for counting as being certain things, of course: I cannot be a philosophy professor at Rice University unless I am hired to be one, hold classes at the appropriate time, and so on. But these – and other, tacit yet no less public – norms do not define what it means to be a philosophy professor, and in acting for the sake of the latter, I can bump up against these public norms. I can decide that giving grades does not belong to what I should do as a teacher, that it is not something that conduces to what I take teaching to mean, and while this may get me fired, it does not mean that I have failed at being a teacher. Trying to be a teacher is nothing but acting in light of what I take to be best in the matter of teaching, and this “best” is nowhere anchored in “being”; it is always at issue in my comportment. So practical identities, no matter how ingrained in my “character” they may seem, are ultimately optional.
For Heidegger, then, one always finds oneself in a “factical” situation in which certain conditions obtain: institutional norms, social practices, and stereotypical roles, but also one’s psychological makeup, drives, inclinations, history, and the like. Though I have no power over what is given in this way, I am not inert in the face of such things; I can choose how I take them up. I stand before them as “possibilities” for behaving in certain ways in light of what I think is best in going on for the sake of being what I am trying to be. As Heidegger puts this point, Dasein “never has power over its being from the ground up,” but it is in such a way that it must “take over being a ground [Grund, reason]” (Reference HeideggerSZ 284).
On Heidegger’s phenomenological account, then, Dasein is, “ontologically,” a response to a normative claim: “you must take over being a ground.”Footnote 5 To act for the sake of being a teacher, for example, is to acknowledge that one is responsible for what one takes to be best in how one goes on. But this means that to take over being a ground is just to take up one’s factical grounds as possible – but only possible – justifying reasons. And because Dasein is essentially “being-with-others” (Reference HeideggerSZ 125), being responsible for reasons is also being answerable (verantwortlich) to others who might challenge one’s reasons. Understood ontologically, then, reasoning is originally reason-giving: a practice that is not definitive of who we are (because it does not derive from a “faculty” we rational animals intrinsically possess) but is not optional for us either (because it arises from a demand to which we, as care, must respond, since what we “are” is precisely at issue).
Heidegger’s phenomenology thus provides a normativity-first account of reason, and if logic is constitutive of the practice of reason-giving, then logic is not optional; logical laws are descriptively constitutive of reasoning, but their normative significance does not derive from “widespread background norms about the relations between belief, reasoning, and truth” (Russell Reference Russell2020, 387). Rather, logic’s normativity follows from the ontological ground of reasoning. It “follows,” not as a logical entailment, a truth-functional consequence of statements about Dasein, but phenomenologically, as an “intentional implication” of our ontological condition as beings who must take over being a ground (reason).Footnote 6 There “is” reason only because we are such that we cannot “be” without acting in the evaluatively normative light of what is best in regard to what we do and say. And this means that reason-giving, giving an “account” of oneself (justification, Rechtgebung), is normatively demanded of a being defined as care (Reference McNeilGA9, 169–170/130–131).
1.5 The Normative Ground of Truth
Following this brief sketch of the phenomenology of comportment and its ontological significance, we may return, more specifically, to the question of how the assertion and its “truth or falsity” should be understood. If the truth of “A is b” is not a property of the sentence but is inseparable from the experience of A as b, what does propositional truth look like in the context of assertive comportment, being a truth-teller?
If we start with the traditional idea that truth is the “conformity [adaequatio, homoiosis] of thinking to the thing that is thought,” our question is now: “what grounds the possibility of conforming to something?” (GA29/30, 497). “Conforming” is a success-term and so belongs within the normative realm, that is, it is intelligible only on the basis of a kind of freedom that can acknowledge norms and bind itself to them. In truth-telling, the thing itself is what provides the measure of what is said, so I must be able to bind myself to the thing as it is. Heidegger claims that “we are never able to say what it is about beings that binds us” because being an “object” (Gegenstand) is normatively indifferent: “not all ‘standing-opposite’ necessarily entails binding.” Thus, “we cannot explain this binding character in terms of objectivity, but vice versa” (GA29/30, 525). Doing so requires a normativity-first approach to objectivity, and so also to propositional truth. The binding character of whatever it is that we encounter and conform our assertion to must become phenomenologically perspicuous in its possibility.
In Fundamental Concepts of Metaphysics Heidegger identifies comportment with noein, a “pre-logical openness” or manifestness of beings “as such” (i.e., as being) and “as a whole.” This is no mere “givenness” but what Heidegger calls world, the normatively structured space of meaning in which beings are encountered as being something (GA29/30, 495–496). Logos, in contrast, and so assertion, is a secondary “manifestness,” one that stands within the possibility of being true or false, measured against the entity to which it refers. But how does the entity become a measure?
As we have seen, every comportment is already a “letting oneself be bound” by that for the sake of which one acts (GA29/30, 496), and in such “letting,” beings as such and as a whole are at issue – disclosed, manifest – in some distinctively meaningful way. For instance, when I act for the sake of being a teacher, when I try to be one, I let myself be bound by norms that pertain to the meaning of teaching, I commit myself to – care about – a measure that is never simply given but remains at issue in my acting. This commitment allows me to apprehend (noein, Ver-nehmen) the world of teaching in which things show up as relevant or irrelevant for the comportment of teaching. In commitment, then, I give myself the measure, measure myself by what I take to be best in the matter of teaching, and this “giving of measure” is eo ipso “transferred to beings in advance in accordance with” my comportment, “so that conformity or non-conformity is regulated by beings” (GA29/30, 496–497). In short, my success or failure at being a teacher cannot be measured if I am indifferent to how things are in the world of teaching.
For instance, from an “objective” point of view a blackboard’s placement in a classroom is normatively indifferent. But if I am trying to be a teacher, I might experience it as badly positioned (A as b), as other than it should be for teaching. Though what allows the blackboard to be experienced as badly positioned is my commitment to teaching, I have “transferred” this “binding character” to the blackboard, which thereby binds me because it belongs to the success-conditions of my comportment: I cannot be a teacher if I ignore the way blackboards are supposed to be in the context of teaching, just as I cannot be a chess player without responding to how a knight is supposed to be moved.Footnote 7
If I find that I need help in moving the blackboard to a better position, I might say to someone, assert, “The board is badly positioned.” I exhibit, determine, and communicate something the measure of whose success (truth or falsity) lies in beings as they are. This is because assertion, as we saw, is assertive comportment, acting for the sake of being a truth-teller, and, as with any comportment, I can only count as so acting if I am not indifferent to how things are. An assertion’s being true or false, then, has the same ontological structure as the blackboard’s being well or badly positioned: The assertion “The board is badly positioned” is true (i.e., is as it should be within the comportment of truth-telling) if (as a teacher) I have encountered the blackboard as failing to be as it should be, and (as a truth-teller), I have succeeded in exhibiting, determining, and communicating this fact. Neither the blackboard nor the expressed assertion (sentence) has any determinate meaning when abstracted from this ontological ground: the comportmental “ap-prehending” (noein) of a normative space, or world, in which things (including sentences) can show up as succeeding or failing at what they are supposed to be. Further, since assertive comportment is a derivative mode of disclosing – that is, since the comportment of being a truth-teller attains its meaning only within some other comportment or practical identity – the “truth” of the sentence in which it is expressed is not only not a property of that sentence; it can get no purchase at all on the sentence in the absence of some other comportment. For instance, the assertion can predicate being badly positioned of the blackboard only if something like the board’s optimal “positioning” is at issue within a context or “world” – here, the world of teaching.
But a major worry remains: It might seem that grounding the normativity of meaning (and so truth as an evaluative norm of assertions) in optional comportments or practical identities leads to a conception of logic as constitutive of a game that we can opt into or out of. But here we see how Heidegger’s ontology allows for a striking reversal of roles. In cases where the comportments on which being a truth-teller depends are contingent practical identities, logic’s normative grip on assertions is likewise contingent; but its normative grip is not contingent when it belongs constitutively to what Heidegger calls our “fundamental” comportment. Such comportment is not optional for us because it is the ontological basis for all comportments: “being free in an originary sense” (GA29/30, 497) is “responding to” the demand to take over being a ground.Footnote 8 No matter what I try to be, taking over being a ground means acknowledging responsibility for the normative force of what I take to be best in trying to be it. Only so can anything become “my” reason. And if Dasein is Mitsein, my responsibility for reasons is answerability to others. Thus reason, as reason-giving (Rechtgebung), is equally fundamental and so not optional for us. For Heidegger, this fundamental comportment, including the demand for reason-giving, is what originally “bears” truth.
Now, reason-giving is a practice in which assertions play a role and, like all practices, it rests on constitutive norms. Thus, no appeal to “widespread background norms about the relations between belief, reasoning, and truth” (Russell Reference Russell2020, 387) is necessary to explain how logic gets a normative grip on reasoning. Logic codifies the norms constitutive of the ontologically demanded “game” of giving and asking for reasons. To talk of the “normativity” of logic is not to say that logic tells us how we “ought” to think, revise our beliefs, or reason. Logic’s “intrinsic” normativity reflects the ontological fact that being answerable is demanded of us and that reason-giving is our response to that demand, a practice for which logic is constitutively normative. Further, because the comportment of truth-telling depends on the fundamental comportment of reason-giving – that is, on the way assertions are to be deployed within the practice of reason-giving – “truth” can, within the limits of assertive reduction to the extant, be “predicated” as a “property” of those assertions, and sentential logic can be formalized into a calculus of truth-preserving entailments.
So, we should agree with Russell that logical laws are descriptive of how truth is preserved in reasoning, but the “truth” in question is not what logic takes it to be. Logical laws do say something about truth, but in treating truth as a property of sentences, they do not say enough about it, since they capture only how truth appears in the dependent (and contingent) comportment of being a truth-teller. But logic’s normative grip does not derive from this contingent comportment. It is intrinsic to the fundamental comportment of reason-giving. On Heidegger’s phenomenological approach, then, logic formalizes the constitutive norms of a practice that answers to what we owe one another, and in which assertions are not first of all truth-bearers but reasons given to others that enable a shared comportment toward something at issue as I account for myself in a dialogical context aiming at mutual understanding.
1.6 Logic, Science, Philosophy
With this result in hand, I conclude with some brief remarks on the implications of Heidegger’s account of the normativity of logic for the three “motivations” Russell provides for thinking that logic is normative.
Beginning with the argument from error (E), if Heidegger holds that logic is constitutively normative for reasoning, then errors in reasoning are either “local” mistakes arising from various kinds of inattention or “global” enough to indicate that the individual does not count as a reasoner at all. But, as we have seen, logic can also be taken to be purely descriptive if one accepts the ontological reduction to the extant that takes place in asserting a sentence. One is then free to abstract sentences from their phenomenological ground and treat them as elements of a system in which “reasoning” in the phenomenological sense does not take place at all. What could motivate such a view of logic?
The descriptivist position might be tempting if one focused on a practice whose constitutive norms seem to align entirely with those of being a truth-teller, that is, if truth-telling were, contrary to fact, not a dependent comportment. And there is indeed something like such a practice – namely, science. According to Heidegger, science is a comportment, “a freely chosen stance of human existence” whose constitutive norm is “to give the matter itself explicitly and solely the first and last word” (Reference McNeilGA9, 104/83). This “submission to beings themselves” has a tendency to conceal science’s status as a comportment, making it easy to embrace the abstraction in which truth appears as a property of the sentences science advances and organizes “theoretically” into a “truth preserving” whole. Such organizing, a reconstruction of knowledge according to logical entailment relations, anticipates the fully developed system of “already known knowledge.” Pursuing scientific knowledge is, of course, a practice, but the proleptic reconstruction of such knowledge is not. It is an “ideal construct” that tells us something about truth – namely, that there can be no violations of logical laws because such laws just describe how truth is preserved in sentential form; they have nothing to do with application. The only price science pays for its self-effacement as a comportment, its “submission to beings,” is to “stray into the legitimate task of grasping the extant in its essential unintelligibility” (Reference HeideggerSZ 153; see Crowell Reference Crowell, Adams and Browning2017).
In Being and Time, Heidegger’s remarks on epistemology suggest that he leaves room for such a view. For instance, in discussing the ontological realism–idealism debate, Heidegger rejects the epistemological approach that begins with developed views on scientific knowledge, while admitting that this approach contains “a grain of genuine inquiry” and has “not gone so very far off epistemologically” (Reference HeideggerSZ 207). Nevertheless, even if the “epistemic” concern with truth as a property of sentences has a certain legitimacy and permits treating logic descriptively as the organizing form of science as “already known knowledge,” it remains an ontological abstraction and cannot fully address the question of the normativity of logic.
This brings us to the argument from normative consequences (NC), which holds that logical validity is intrinsically normative thanks to a direct consequence it has for what we ought to believe. Here Heidegger’s phenomenological approach makes common cause with Gilbert Harman: Logic has no direct normative consequences for belief-formation and revision because belief is independently beholden to another norm, that of evidence. Husserl formulated this as the “principle of all principles: that each intuition affording [something] in an originary way is a legitimate source of knowledge,” arguing, further, that “no conceivable theory can make us stray” from this principle because any theory can itself “draw its truth only from originary givenness” (Husserl Reference Husserl2014, 43). Despite having some objections to Husserl’s formulation, Heidegger clearly embraces its basic sense: Belief, as cognition, must ultimately be based on “exhibiting [something] directly and demonstrating [Ausweisung] it directly,” where “directly” refers to the kind of originary encounter or “intuition” of the matter in question (Reference HeideggerSZ 35).
So here too we find a qualified agreement with the descriptivist account of logic, but the qualification is important. Evidence and belief-formation presuppose comportments that ground practices, and all such comportments are already normatively constrained by the “fundamental comportment” of taking over being a ground. As we have seen, fundamental comportment phenomenologically entails answerability, such that logic is constitutively normative for the nonoptional practice of reason-giving. Any conflict between reasoning as logical inference and reasoning as appeal to evidence must be adjudicated “pragmatically” through actual argumentation (Habermas Reference Habermas1990, 94).
This, in turn, provides Heidegger’s answer to the argument from demarcation (D). Motivation for conceiving logic as an account of how we “ought” to reason arose from the worry that the descriptivist position could not decide between mutually exclusive accounts of logical entailment, suggesting that logic might be altogether conventional. Russell defended descriptivism by arguing that deciding between competing accounts of entailment is an “epistemic” matter requiring better “semantics” – “more theory and more evidence” (2020, 386). While agreeing that logic is not about how we “ought” to reason, Heidegger shows that the descriptivist position is not the last word on logic’s normativity. If truth is taken to be a property of sentences, and if sentences have meaning only within practices, then the decision about which notion of entailment best models a given practice may indeed be conventional. But if logic is constitutive of the “game” of giving and asking for reasons, then its “domain,” reasoning, is not conventional but is ontologically grounded in something normatively demanded of us as beings who are answerable to others for what we do and say.
Unlike reasoning, philosophy is not something demanded of us by our very being; it is optional. And like all comportments, it stands under the “obligation” to give reasons, to answer for what, if anything, it asserts. However, phenomenological philosophy is not constituted, like science is, by the norm of giving beings the “first and last word.” Rather, it aims at elucidating the being of beings – that is, the meaning thanks to which beings can show up as the beings they are. Philosophy does aim to speak truly, and so the comportment of being a truth-teller belongs to it as it belongs to science. But to tell the truth about being (meaning) is not necessarily to assert a sentence that can be seen as having a property, truth, that is preserved in relations of logical consequence. What looks like such a sentence may, in philosophy, be nothing of the sort. The “yield” of philosophical assertions must be understood from within the project of philosophy. Perhaps the “dependent” comportment of truth-telling ought not always observe logical entailment relations. Perhaps the form that reason-giving takes in philosophy cannot be formalized into a calculus. Indeed, on Heidegger’s account of philosophy, the reduction to the extant (the “specialty of assertion”) is phenomenologically offset by the use of “formal-indicating concepts” (GA29/30, 421–431) that lead the addressee of philosophical sentences back to the evidence-situation that alone provides access to what is exclusively of concern in philosophy: the being of beings (GA61, 157; Crowell Reference Crowell2001, 129–151).
For a time, Heidegger held that phenomenological ontology was “objectification of being” or “transcendental science” (GA24, 465), but he soon abandoned this idea as a holdover from the metaphysical tradition. In later writings Heidegger prefers the term “thinking” for inquiry into being. Thinking is neither cognition nor reasoning but phenomenological Besinnung – sich auf den Sinn einlassen (GA7, 63) – and in that context, what looks like a false assertion (for instance, a metaphor) might nevertheless communicate the shared comportment toward being required of thinking as the project of bringing meaning to light. If thinking is allowed the metaphorical “use” (Davidson Reference Davidson1984, 247) of sentences – and, as Hannah Arendt pointed out, thinking “cannot avoid them” (1978, 112) – this would signal a priority of phenomenology over logic in thinking, not an equal footing. If access to meaning comes not through cognition and reasoning but through thinking, phenomenology as Besinnung, then it may well be that the “reign of logic” in philosophy “disintegrates into the turbulence of a more original questioning” (Reference McNeilGA9, 117/92).