Thick Ethical Concepts — Comprehensive Study Notes

  • Thick concepts overview

  • Thick vs thin concepts: thick = evaluate plus descriptive content; thin = primarily evaluative (e.g., right, good, ought) or descriptive only

  • Thick concepts include virtue/vice terms (generous, selfish), practical terms (shrewd, imprudent), epistemic terms (open-minded, gullible), and aesthetic terms (banal, gracious)

  • Thick terms contrast with thin terms which are often used to express evaluation without substantive non-evaluative description

  • Concepts vs terms: concepts are non-linguistic representations; terms are linguistic items expressing concepts; small capitals denote concepts, italics denote terms

  • Thick concepts are argued to be action-guiding and world-guided; they tie evaluation to world-sensitive description

  • The rise of thick concepts: important for debates on is/ought, fact/value, objectivity, and how evaluative classifications relate to non-evaluative ones

  • Thick concepts appear across ethics, aesthetics, epistemology, and philosophy of science

  • Core questions driving thick concepts

    • The combination question: how do thick terms combine evaluation and non-evaluative description?

    • The location question: is evaluation inherent to thick concepts or a feature of social use?

    • The delineation question: how do thick concepts differ from thin concepts and from other evaluative terms?

  • This notes set surveys major approaches and debates on thick ethics concepts

1. What Are Thick Concepts? Background and Preliminaries

  • Origin of the term “thick concept”: Bernard Williams’s Ethics and the Limits of Philosophy (Williams 1985)

  • Williams borrowed “thick” from thick description (Ryle; Geertz): thick descriptions include intentional, purposive details and cultural significance

  • Thick descriptions need not involve evaluation; thick concepts do involve evaluation plus descriptive content

  • Cultural specificity: thick concepts may vary across societies (e.g., courage, chaste, noble) due to different moral/environmental contexts

  • Early contrasts: thick terms like generous, cruel, tactful appear to carry descriptive content beyond merely good/bad

  • Key questions about thick concepts

    • What counts as descriptive vs. evaluative in thick terms?

    • How can thick terms be understood if their evaluations are context-sensitive or world-guided?

    • Are thick concepts purely evaluative or do they always embed non-evaluative content? Could they be dual-character concepts (content that is both descriptive and evaluative but not reducible to one or the other)?

  • Dual character concepts: some propose a split where a term (e.g., artist) may count as artist in one sense (descriptive) and in another sense (normative or aspirational); this can be viewed as polysemy or as two content-dimensions that contextually weigh differently

  • Shapelessness idea: evaluative content of thick terms may not be capturable by independent non-evaluative content; this helps explain why thick terms resist straightforward decomposition

  • Significance of thick concepts

    • They challenge strict is/ought boundaries and the fact/value dichotomy

    • They factor into debates on objectivity and how evaluative classifications relate to non-evaluative ones

    • They offer a broader framework for understanding ethical reasoning, normative judgments, and culture-specific practices

  • Summary of stance: thick concepts blend evaluation with world- and context-dependent description; their study aims to clarify how this combination works, where evaluation sits relative to content, and how thick terms relate to thin terms

2. Do Thick Concepts Have Distinctive Significance?

  • Two influential lines of argument for distinctive significance

    • The Is-Ought Gap and the Fact-Value Distinction

    • The Anti-Disentangling Argument

  • 2.1 The Is-Ought Gap and the Fact-Value Distinction

    • Is-Ought gap: non-evaluative (descriptive) premises cannot guarantee evaluative conclusions (Hume’s law)

    • Thick terms could threaten this gap if their evaluative content is entailed by non-evaluative premises

    • Foot’s argument: thick terms like “rude” may be analytically entailed by non-evaluative descriptions (x causes offense by indicating lack of respect → x is rude)

    • Counterpoint: this strict entailment leads to controversial implications, especially for objectionable thick terms (e.g., chaste, lewd, selfish)

    • Key inference: if some thick terms semantically entail evaluation, they could undermine the is-ought gap; if not, the gap remains

    • Possible replies to Foot

    • Thick terms may be evaluative in meaning yet not semantically entailing non-evaluative content; evaluation could be conventional or pragmatic rather than semantic

    • Some argue for a distinction between semantic content and conventional or pragmatic use; thick terms could be used evaluatively without semantically entailing non-evaluative content

    • The debate continues on whether there is a clean semantic entailment from non-evaluative descriptions to evaluative conclusions

  • 2.2 The “Anti-Disentangling” Argument

    • Core idea: if there is a clean, co-extensive non-evaluative description D for a thick term T such that D suffices for T, then T would be disentangled into non-evaluable content plus evaluation

    • McDowell’s disentangling argument (anti-disentangling): for some value concepts, there is no non-evaluative description that is co-extensive with the evaluative meaning; master the extension of T without grasping its evaluative orientation would be impossible

    • Consequences depend on assumptions like Descriptive Equivalence (DE): the extension of T is determined solely by non-evaluative content; if DE is true, anti-disentangling supports Inseparability (thick concepts cannot be separated into evaluative + descriptive parts)

    • Rebuttals and complications

    • Separability does not entail Descriptive Equivalence; even if D and E are separable, DE may fail; thick terms could be analyzable as an explanatory relation (e.g., “x is E in virtue of D”) rather than a simple conjunction of D and E

    • Kyle’s Expansion View and Elstein & Hurka patterns show ways to maintain Separability without DE or full reduction

    • Shapelessness and its role: even if anti-disentangling appears strong, the absence of a neutral non-evaluative profile for all thick terms (shapelessness) can undermine the idea that a non-evaluative description could exhaust the term’s extension

  • 2.3 Reflection, Knowledge, and Priority

    • Bernard Williams’s claim: in ethics, reflection can disrupt knowledge; thick concepts may be parochial and tied to particular evaluative outlooks

    • Hypertraditional society thought experiment: a highly reflective agent may come to reject thick concepts by adopting thin concepts, suggesting a potential priority of thin concepts in evaluation and knowledge

    • The debate about whether thick concepts are foundational or derivative: three main positions

    • Thin Centralism: thin concepts are prior to thick; grasp of thick concepts requires grasp of thin (Hare; Blackburn; Elstein & Hurka with embedded evaluation)

    • No Priority: thick and thin are interdependent; neither strictly prior

    • Thick Centralism: thick concepts are prior to thin; thick content shapes or determines how thin concepts apply

    • Consequences for cross-cultural understanding and objectivity: if thick concepts are central to evaluative practice, there may be no single homogeneous evaluative language across cultures or communities; reflection can shift or revise evaluative outlooks

  • Summary: Distinctive significance arguments center on whether thick concepts challenge is/ought and fact/value distinctions, whether they resist reduction, and how reflective practices affect knowledge and evaluation

3. The Combination Question: How Do Thick Concepts Relate Evaluation and Description?

  • Core issue: are thick terms reducible to a non-evaluative description plus an evaluative component, or are they irreducibly thick?

  • Separabilist view (thick terms have separable evaluative and descriptive aspects)

    • Core idea: D (non-evaluative description) and E (evaluative content) are distinct contents; T’s meaning may rely on both

    • Typical approach: E is often taken to be evaluative content (sometimes in truth-conditions) and D provides non-evaluative grounding

    • Problems: if D entails E, we face the is/ought problem and potential Descriptive Equivalence; if D does not entail E, there must be an explanatory link binding them

    • Burton’s idea: thick terms meaning “x is E in virtue of some particular instance of D”; allows context-sensitive reliance on particular features, avoiding full entailment

    • Kyle’s Expansion View: thick terms are expanded contents of thin terms (e.g., admirable = worthy-[of-admiration]); this preserves separability without requiring a full reduction

    • Elstein & Hurka’s two patterns for reductive analysis

    • Pattern 1: Global evaluation governs the whole concept; non-evaluative content specifies broad category (e.g., distributively just means it is good; contains general non-specifiable features X, Y, Z that make distributions good)

    • Pattern 2: Embedded evaluation; there is an internal thin evaluation that is embedded within non-evaluative description (e.g., courageous means good and involves accepting harm for greater goods)

    • Consequences: separability need not imply full reduction; evaluation can drive extension while non-evaluative content grounds application

  • Inseparabilist view (thick concepts are irreducibly thick)

    • Core idea: thick meanings fuse evaluation and descriptive content in a way that cannot be disentangled into independent components

    • Examples: cruel expresses a sui generis evaluative concept

    • Challenge: explain how thick terms are evaluative and descriptive without a simple two-part structure; need a theory of how they are evaluative yet descriptively informative in a non-reducible way

    • Relation to thin concepts: thick and thin evaluative meanings may share practical relevance, but thick terms often carry context-dependent point or reason for action

    • Metaphysical and epistemological issues: how thick terms relate to reasons, motivation, and objectivity; debate over whether thick terms are inherently evaluative

  • The debate over separability vs inseparability intersects with broader questions about centralism, shapelessness, and the nature of evaluation

4. The Location Question: How Are Thick Concepts Evaluative?

  • Central claim: many philosophers agree T-utterances convey evaluation; the question is where the evaluation sits: semantic content, pragmatic not-at-issue content, or presupposed content

  • 4.1 The Semantic View (evaluation as semantically expressed content)

    • Core claim: global evaluation is a primary semantic content of sentences involving thick terms; evaluation is semantically expressed

    • Data supporting view: examples where denying the evaluation in a thick-terms sentence leads to infelicity or contradiction (e.g., cruel is bad in a particular way; “It was courageous of Sue to stand up to that racist on the bus” vs. “courageous—and not good in any way for it”)

    • Cognitivist interpretation: evaluation is a truth-conditional content; expressivists see it as speaker attitude rather than truth-conditional content; but many acknowledge thick terms may vary in valence across contexts

    • Objections and complications

    • Contextual variability in valence: thick terms can convey positive in some contexts and negative in others (e.g., frugal; cruel)

    • Such variability could be seen as a challenge to a strictly semantic view; proponents suggest this is compatible with the Semantic View if we allow context-dependent evaluation or pragmatic meaning

    • Projection data: evaluations associated with thick terms often project under negation, questions, or embedding (e.g., chaste in projection contexts); this challenges a pure semantic account where evaluation is strictly embedded in truth-conditions

    • Denial/defeasibility data: direct denial of evaluation after assertive T-utterances is not straightforward; thick evaluations can be suspended or canceled in nuanced ways (e.g., “Isolde is chaste” vs. “Isolde is chaste, but not good in any way”)

    • Possible reconciliations

    • Not-At-Issue content: evaluation may be a general implication of T-utterances that projects but is not the primary asserted content; not-at-issue content explains projection and defeasibility without abandoning semantic evaluation

    • Gricean pragmatic explanations: conversational implicatures may carry evaluative import; however, these often fail to explain all defeasibility data and projection

  • 4.2 Secondary Content Views (evaluative content is conveyed via secondary mechanisms, not semantically expressed)

    • Conventional evaluation views

    • Evaluation conveyed via presupposition or conventional implicature

    • Presupposition example: “I don’t regret being a member of the Communist party” presupposes membership; evaluative presuppositions could present the target as bad in a certain way or express speaker attitude

    • Problems: presuppositions project and may be defeasible; some evaluative presuppositions are not easily captured by semantic content; accommodation issues arise when presuppositions are challenged by listeners

    • Conventional implicature: e.g., “Shaq is tall but agile” suggests agility is surprising given tallness; slippery to separate from semantic content; some argue evaluative content may be detachable, others deny this

    • Pragmatic views (conversational implicature and other pragmatic mechanisms)

    • Gricean pragmatic implicature: evaluation is conveyed by implicature rather than semantic content; usually cancellable in direct negation (contrasts with thick terms that tend to resist direct cancellation)

    • Notable problem: thick-utterance defeasibility and cancellation patterns often do not align neatly with standard conversational implicature (e.g., “Isolde is chaste, but I don’t mean to imply she’s better in any way” is awkward; defeasibility patterns differ from classic implicature)

    • Not-At-Issue View (Väyrynen)

    • Global evaluations are not typically at issue; they project and are defeasible but are not part of the literal primary content; evaluation arises as a background not-at-issue content

    • This view explains projection and defeasibility without requiring semantic expressivity of evaluation; thick terms may still be evaluative, but evaluation is not semantically encoded in a straightforward way

    • Overall assessment of Secondary Content Views

    • Pros: capture projection and defeasibility data; flexible to context; can accommodate cross-context variability

    • Cons: may not predict all data; some objections to thick terms’ evaluative content being purely secondary; still under debate how best to combine semantic vs pragmatic mechanisms

5. The Delineation Question: How Do Thick and Thin Concepts Differ?

  • Core difference: how thick vs thin differ non-evaluative content

  • Two broad options for difference

    • Difference in kind: thick concepts involve both evaluative and non-evaluative content that is essential; thin evaluative concepts are purely evaluative or less descriptively rich

    • Difference in degree: even thin concepts may have some non-evaluative thickness; thick and thin lie on a continuum of descriptive thickness

  • Williams’s view (difference in kind): thin concepts are purely action-guiding, thick concepts are world-guided as well; this implies a robust non-evaluative content for thick concepts

    • Problems: some thin concepts (ought, right) may have non-evaluative content; the distinction is not theory-neutral

  • Scheffler’s continuum view: many evaluative terms lie between thick and thin; not all terms fit neatly into a binary classification; some terms may be thick in some respects and thin in others

    • Consequence: a binary thick-thin split might be too crude; a graded approach could better capture evaluative terms’ thickness

  • Centralism vs No Centralism (relation between thick and thin)

    • Thin Centralism: thick concepts can be analyzed in terms of a few thin evaluative concepts plus non-evaluative content; thickness is reducible/derivable in practice

    • Thick Centralism: thick concepts are conceptually prior or irreducible; they cannot be fully captured by thin concepts

    • No Priority: neither term type is strictly prior; both contribute to evaluative thought and discourse

  • Hare’s view (degrees of attachment)

    • Distinguishes primary vs secondary evaluative content in terms of how firmly attached they are to the term’s usage

    • For thin terms, evaluative meaning may be more firmly attached to action guidance; for thick terms, non-evaluative meaning may be more firmly attached; this perspective can accommodate a degree-based understanding without commitment to a strict semantic vs pragmatic split

  • Takeaway: the thick-thin distinction invites multiple theoretical options, including difference in kind, difference in degree, or a hybrid; there is no single neutral resolution; debates continue about how best to characterize their relation and what this means for metaethics, epistemology, and philosophy of language

6. Thick Concepts Outside of Ethics

  • Thick concepts extend beyond ethics into other domains

    • Aesthetics: thick aesthetic concepts (garish, delicate, balanced) vs thin aesthetic terms (beautiful, ugly); debates about whether aesthetics can be reduced to thin concepts or involve thick evaluative content

    • Epistemology: thick epistemic concepts (gullible, open-minded, quick to jump to conclusions) and virtue epistemology; questions about whether knowledge is itself thick and how it interacts with evaluative content

    • Law: legal concepts like crime and inheritance may be thick, combining descriptive institutional facts with evaluative endorsement; suggests law’s nature may be thick in a way that informs philosophy of law

    • Logic: some argue that logical validity is a thick concept

    • Science: concepts like wellbeing, resilience, biodiversity, sustainability; discussion of whether scientific thick concepts reflect evaluative judgments or aim for objective descriptions

    • Applied philosophy: thick concepts appear in risk theory (risk and safety), environmental ethics (ecological integrity), and other applied areas; thick concepts can ground normative standards and policy implications

  • Cross-disciplinary relevance and implications

    • Thick concepts can illuminate how normative language functions in scientific, legal, and social domains

    • They raise methodological questions about measurement, value-laden discourse, and objectivity across disciplines

  • Summary of outside-ethics significance

    • Thick concepts challenge simplistic separations between fact and value in multiple domains

    • They invite cross-disciplinary methodological approaches to understand how evaluative content is integrated with descriptive content

7. Additional Notes and Connections

  • Dual character concepts: similar to thick concepts but with potentially independent descriptive and evaluative dimensions; debate about whether dual-character concepts are normative/evaluative in the same way as thick concepts

  • Shapelessness: thick concepts may not map neatly to independently intelligible non-evaluative descriptions; this supports the view that evaluation is not neatly separable from description

  • Notable debates and references

    • Is-ought gap, Foot’s critique of entailed evaluation, and the anti-disentangling argument

    • McDowell’s anti-disentangling argument and its implications for Separability vs Inseparability

    • Expansion View (Kyle): thick concepts as expanded contents of thin terms; preserves separability without full reduction

    • Semantic View vs Secondary Content Views: ongoing debate about whether thick concepts semantically express evaluation or convey it via pragmatic/secondary means

    • Not-At-Issue View (Väyrynen): evaluation as not-at-issue content that projects and is defeasible

  • Practical and ethical implications

    • Thick concepts influence normative criticism, value judgments, and debates about objectivity and justification in ethics

    • They also shape cross-cultural understanding and the interpretation of moral language in diverse communities

    • Ethical reflection and knowledge: thick concepts complicate the idea that ethical knowledge is straightforwardly gained through pure deduction or universal principles; reflection can alter or replace traditional thick evaluative frameworks

8. Summary of Core Concepts

  • Thick vs thin terms and concepts

    • Thick terms combine evaluative content with non-evaluative descriptive content

    • Thin terms are primarily evaluative and less tied to descriptive content

  • Combination question

    • Do thick terms’ evaluative and descriptive parts constitute separate contents or are they irreducibly fused?

  • Location question

    • Is evaluation semantically expressed or is it conveyed pragmatically or via secondary content (presupposition, implicature, etc.)?

  • Delineation question

    • Do thick concepts differ from thin concepts in kind, degree, or via a mixed model?

  • Distinctive significance

    • Thick concepts challenge neat fact/value distinctions and influence debates on ethical reasoning, objectivity, and epistemology

  • Outside ethics

    • Thick concepts are increasingly relevant in aesthetics, epistemology, law, science, and applied philosophy; they shape cross-disciplinary discussions about value-laden language and evaluation

Note: All mathematical notation in this note is provided using LaTeX syntax where appropriate. For example:

  • If explicit non-evaluative content D and evaluative content E are connected, one might denote their relation as D<br>ightarrowED <br>ightarrow E or D<br>ot<br>ightarrowED <br>ot<br>ightarrow E depending on whether D entails E.

  • Expanded contents and embedded evaluations can be represented as:

    • extadmirable=extworthyextofadmirationext{admirable} = ext{worthy}- ext{of-admiration}

    • extcourageousext(embedded)<br>ightarrowextgoodext{courageous} ext{(embedded)} <br>ightarrow ext{good}

  • Notation for impact of evaluation on action guidance: extEactsasareasonext{E acts as a reason} in some theories; others treat E as a property of usage or discourse rather than a standalone truth-conditional content.

Title: Thick Ethical Concepts — Comprehensive Study Notes