Relatively True or Truly Relative? A critical summary of "On Rightness of Rendering" by Nelson Goodman
Art, literature and even scientific papers can be interpreted in an infinite number of ways. So what is it that makes one particular interpretation of a rendering right?
In a world of an infinite number of possible interpretations, what is it that makes one particular interpretation of a given “rendering” correct? By what standard should rightness be measured? Truth? Validity? Accuracy? Or perhaps a combination of both that includes truth but extends to other criteria that “compete with or replace truth under certain conditions”?
This is the position Nelson Goodman bats for in his essay On Rightness of Rendering and my aim is to explain and summarise how he arrives there.
On Rightness of Rendering explores what it is that makes a “rendering” – i.e. a symbolic representation of a world in language, visual, scientific or musical in form – “right” – i.e. what interpretation amongst a seemingly infinite list of possible ones is, to crudely simplify Nelson Goodman’s position ‘most correct’.
Goodman examines in detail how the relationship between ‘truth’[1] and communication of that truth via language, is affected by symbolic representations that we make. He applies analytic philosophy and logic to the domain of language and aesthetics to propose “rightness” as a criterion for evaluating representations of worlds, as an alternative to the conventional binary notion of truth[2] which he argues is limited to declarative sentences. Goodman aims to identify standards of acceptability for evaluating the value of non-declarative renderings, noting that the way we perceive the world is not restricted to literal deductive or inductive statements about the way the world is[3] but also includes literary, musical and aesthetic statements whose truth or correctness is more difficult to determine.
He does this by evaluating the efficacy of the following acceptability standards: truth, validity and finally “rightness” as acceptability standards for renderings – which to repeat Goodman’s definition again, means representations of worlds both scientific and aesthetic. The latter, Goodman suggests, is particularly prone to the post-modern observation alluded to earlier – multiple conflicting interpretations exist, all of which appear equally rationally acceptable. These kinds of statements – i.e., statements for which multiple equally acceptable interpretations exist – do not conform to the traditional binary notion of being either true or false. For these such cases Goodman calls “non-declarative renderings”, he proposes “rightness” as a more appropriate standard of acceptability that includes truth, but extends beyond it to other standards that “may supplement, compete with or even replace truth” under certain conditions.
Goodman explains the reason for his title. He uses the term “rendering” to distinguish the rightness being discussed from moral rightness and defines it as “all the ways of making and presenting worlds” including in that definition both aesthetic and scientific representations of worlds.
1. Truth as an acceptability standard
Goodman considers the difficulties with establishing truth from statements, using the examples like "the earth stands still" and "the earth dances the role of Petroushka" to illustrate that the truth of the sentences are relative to the interpretive framework[4] used. In they were answers on a true-or-false test, both would be marked false, but Goodman explains that the truth of these propositions shift in a different condition. For example, he considers the possibility that propositonal truth shifts relative to specific frameworks—Ptolemaic astronomy or Stravinsky’s ballet—(e.g., "In the Ptolemaic system, the earth stands still"). The upshot is that the truth of a statement and its framework are interdependent: one might fixate on the statement whilst ignoring it’s framework, or vice versa and in doing so be lead to error. For example, a guard ordered to shoot moving prisoners, who justifies killing immobile captives by citing their motion relative to the earth’s axis, illustrates how a proposition that is anti-correlated to its framework can lead to error. Goodman contrasts two kinds of frameworks:
1. Compatible— like the Ptolemaic model of the universe and a certain style of ballet choreography—because they can be translated into each other or used side by side without conflict.
2. Conflicting Frameworks— like two historians giving different facts about the same event. For example, if Herodotus says Spartan kings had two votes and Thucydides says they had one, both propositions cannot be literally true simultaneously. But if we prefix the propositions with “According to Herodotus…” or “According to Thucydides…”, we can see both statements as true within their own frameworks, independent of which one is historically accurate[5].
While relativisation resolves ambiguity in some cases, it can amplify ambiguity in other cases. Goodman explores a third case—perception versus physics regarding a moving spot—where "the spot moved" (perception) and "no spots moved" (physics) seem at odds. He rejects reductionist views (e.g., translating one into the other) and absolutist stances (e.g., one being true, the other false), proposing instead that both may be true within their respective domains, creating distinct "worlds." This challenges the idea of a singular reality, suggesting that irreconcilable truths necessitate a shift from truth to rightness as the evaluative standard.
Validity
Goodman then examines deductive validity as a standard of acceptability. Unlike truth, validity hinges on conformity to inference rules—a valid argument can have false premises and conclusions yet maintain "right" structure. However, a valid inference from true premises must yield true conclusions. Inductive validity, by contrast, is further removed. A valid inductive argument from true premises (e.g., "all observed emeralds are green") need not guarantee a true conclusion ("all emeralds are green"), especially if negative evidence is omitted or categories are unprojectable (e.g., "grue"—green before a date, blue after).
Rightness here requires comprehensive evidence and projectable predicates, rooted in entrenchment (habitual use) and modified by innovation (new insights). This leads to rightness of categorization, where efficacy, not truth, governs, as categories like "green" versus "grue" lack truth value but differ in utility for understanding.
Goodman then applies this to pictorial representation, refuting the simplistic view that what makes a picture “right” is resemblance. For Goodman, a picture’s rightness depends on its system—perspective, colour conventions, or distortions like complementary colours can be "right" within their frameworks, much like descriptions of planetary motion vary by system (e.g., heliocentric vs. geocentric). Realism splits into two senses: adherence to entrenched norms (e.g., a Dürer painting over a Cézanne in Western tradition) and revelation of new aspects (e.g., a painter uncovering overlooked realities). Both reflect the interplay of habit and novelty seen in induction and categorization.
For abstract works—lacking denotation or depiction—rightness involves design and exemplification, where a work presents properties (e.g., patterns, emotions) rather than describing them (e.g., a painting of a landscape, or still life drawings). Goodman uses the analogy of a "fair sample" (e.g., a cloth swatch or seed mixture) to explore this. A swatch might be unfair if it is too small to show a pattern or cut misleadingly; a seed sample might fail if its proportions don’t match the whole. Fairness isn’t about exact replication but projectability—reliably extending properties from sample to whole based on entrenchment (familiarity) and adaptability (innovation). In art, an abstract work is right if its exemplified features (e.g., rhythm, form) can be projected onto a world, shaping perception and understanding.
Conclusion: Rightness
Goodman unites these standards of acceptability under a general notion of rightness, distinct from yet akin to truth, driven by two forces: inertia (entrenchment, habit) and innovation (inquiry, novelty). This applies across science (e.g., preferring elegant laws), art (e.g., valuing new perspectives), and perception (e.g., adapting frameworks). He dismisses beauty or pleasure as primary criteria, focusing on efficacy in world-making—organizing experience into coherent, insightful versions. Unlike truth, rightness accommodates variability without collapsing into subjectivity, mirroring scientific standards in its balance of constancy and change.
Goodman concludes that knowledge transcends mere truth-seeking, encompassing the determination of rightness across all renderings, even where truth is irrelevant. This expansive view—embracing fabrication of facts, invention of schemata, and discernment of affinities—requires further exploration, particularly for non-verbal, non-literal versions. He leaves the task open-ended, acknowledging the need for detailed criteria and deeper investigation into how rightness shapes understanding.
Critique
Goodman’s idea of conflicting frameworks is fascinating, but appears to my eye, to equivocate on truth value – i.e. he could be read to be saying that both Herodotus and Thucydides statements are true in the sense that they are accurately describing history. Goodman appears to be saying that we build worlds by building an accurate representation of the way that we perceive the world to be – resulting in two irreconcilably conflicting but paradoxically, equally true accounts. I agree with the first idea, but my worry is that Goodman’s second idea (that the two accounts are equally true) fails to properly distinguish between the subjective (ie. Thucydides perception of historic events) and objective (ie. the historic events that in fact took place). Additionally, Goodman overlooks an important part of his own rightness standard – that is that “truth” (independent of validity, frameworks or other conditions that might replace it) is deeply important as regard the way that the actual world is… independent of our cognitive constructions of it.
[1] Or the “historical” framework
[2] Goodman doesn’t explicitly define what a framework is, probably because it’s fairly obvious but I think that by framework Goodman refers to a way or “format” of rendering. Ie. a way of representing the world.
[3] Used in quotation marks because this is of course a slightly crude representation and not quite how Goodman defines it later on as we will see
[4] That is a proposition is either true, or it is false
[5] For example, statements like scientific or mathematical propositions