Enrahonar. Quaderns de Filosofia 49, 2012 155-174 RESSENYES Shottenkirk, Dena (2009) Nominalism and Its Aftermath: The Philosophy of Nelson Goodman Dordrecht: Springer, IX, 170 p. Nominalism Unscathed ognizes, themes from all three interweave Nelson Goodman is widely considered Goodman’s work, so it would be a mis- one of the foremost philosophers of the take to end the discussion there. 20th century. He made significant con- To understand the motivation be- tributions to metaphysics, epistemology hind Goodman’s nominalism, some and the philosophy of art. Standardly his background may be helpful. According works are considered in isolation, and are to naïve set theory, any collection of not seen as resting upon each other in items constitutes a set. Unfortunately, any significant way. In Nominalism and this gives rise to the Russell paradox. The its Aftermath: the Philosophy of Nelson set of all sets that do not contain them- Goodman, Dena Shottenkirk (Lecturer at selves contains itself if and only if it does Brooklyn College in New York) attempts not contain itself. Although intuitively to unite Goodman’s philosophical enter- plausible, it is inconsistent. Further, be- prises under the nominalist framework cause arithmetic reduces to set theory, if he worked in. This is an insightful and naïve set theory is set theory, mathemat- promising line of inquiry. However, the ics as a whole rests on shaky ground. book falls short of its goal due to miscon- Thus several mathematicians including ceptions of Goodman’s philosophical Russell developed amended versions of positions. set theory. Russell’s ramified theory The work is composed of three main of types constrains set membership in a sections that focus on Goodman’s meta- manner that he describes as «perverse» physics, epistemology and philosophy of (Bertrand Russell 1908). Every other art. Each section draws heavily on the non-naïve set theory contains elements conclusions of its predecessors. Given the that are equally unintuitive. All are des- chronological order of Goodman’s major perate expedients to restore consistency. publications (The Structure of Appearance Nelson Goodman took a much bold- (1951), Fact, Fiction and Forecast (1955), er course. He denied the existence of sets, and Languages of Art (1968)), this is restricting his ontology to individuals. plausible. However, as Shottenkirk rec- The two guiding principles governing his 156 Enrahonar. Quaderns de Filosofia 49, 2012 Ressenyes nominalism are ‘entities are identical just that «philosophy is not about ‘the world’ in case their parts are identical’ and ‘there but about the language through which we is no entity without an identity’. The first see the world, and the latter is primarily principle excludes set theory, because ascertained through science» (68). This is many different sets have exactly the same wrong on multiple counts. It is at best elements. The second excludes intension- misleading to say that Goodman thinks al entities such as properties, because they that philosophy is not about the world. lack determinate identity conditions. He holds that symbols refer to items in Anything that satisfies the two conditions the world. The word ‘cat’ and a picture qualifies as an individual — so it is pos- of a cat is about a cat, not about the way sible for abstract, concrete, mental and we represent cats. The word ‘individual’ physical things to be individuals. Good- is about an individual, not about the way man devoted much of his early career to we represent individuals. Goodman does determining what could be accomplished not hold that understanding is couched with such a limited ontological frame- exclusively in language. In Languages of work. He and Henry Leonard developed Art he claims that works of art function a version of mereology — the calculus of in non-linguistic symbol-systems which individuals (Nelson Goodman and enable us to understand things in non- Henry Leonard 1940). In ‘Steps toward linguistic ways. He denies that under- a Constructive Nominalism’ (1947) standing is primarily achieved through Goodman and W. V. Quine investigated science. He insists that «the arts must be how much of mathematics could be sal- taken no less seriously than the sciences vaged in an ontology without sets. as modes of discovery, creation and en- Philosophers standardly assume that largement of knowledge in the broad so drastically restricting metaphysical re- sense of advancement of the understand- sources must yield an impoverished phi- ing» (WW, 102). losophy. However, in The Structure of One of the most glaring mistakes in Appearance, Goodman shows that es- Nominalism and its Aftermath concerns chewing sets, properties and natural the presentation of the grue paradox. kinds can be liberating. We can acknowl- Goodman is well known for demonstrat- edge that members of any extension have ing that we do not have an adequate con- something in common — i.e., the mem- ception of inductive validity. Previously, bership in that extension. Atypical kinds many thinkers accepted the principle may be interesting and informative. that if one has observed many A’s in the There is no incentive to think that there past and they have all been B’s and no is a single, basic way the world is. In The A’s that are not B’s, one has reason to Structure of Appearance Goodman devel- believe that unobserved A’s are also B’s. ops a phenomenalist system to illustrate Then Goodman introduced ‘grue’. the power of his form of nominalism. Something is grue just in case it is ob- But he does not claim that it does or served before time t and found to be should underwrite physicalism. There are green, or not observed before time t and many ways things can be organized. Phe- is blue. If time t is in the future, all ob- nomenalism is a way of structuring ap- served emeralds have the property of pearances. being grue — for all emeralds have been The Structure of Appearance is a work observed before time t and have been in formal philosophy that is heavily influ- found to be green. Given the inductive enced by Carnap’s Aufbau (1928). Ob- principle set out above, we have induc- serving parallels with logical positivism, tive evidence that all emeralds are grue. Shottenkirk says that Goodman claims But the same evidence supports the con- Ressenyes Enrahonar. Quaderns de Filosofia 49, 2012 157 clusion that all emeralds are green. So by tions and coextensive terms in projec- the above principle, our observations tions by speakers in any community. provide reason to believe that an emerald Further, entrenchment is not sufficient not observed before t is both green and for projectibility. A projectible hypoth- blue. This is preposterous. Shottenkirk esis must also have known positive in- mischaracterizes the problem. She states: stances, no known negative instances, «Since evidence statements regarding all and some undetermined instances. the emeralds before t confirm that they Whether or not a given predicate like are all grue, they unfortunately confirm, «green» was projected in the past is de- in turn, both of the competing hypoth- pendent on the ways various linguistic eses, e.g., 1) that all emeralds are green communities used words. But whether a and 2) that all emeralds are blue» (84). predicate is projectible is not a matter of This is incorrect. The observation of community consensus. green emeralds does not confirm the hy- Shottenkirk intimates that the grue pothesis that all emeralds are blue — in- paradox is problematic for Goodman be- deed, it disproves it. If all that is required cause his nominalist framework does not for confirmation is conformity to the allow natural kinds. If this were so, the evidence, the observation confirms that paradox would not have had the monu- emeralds that have not been observed mental impact on epistemology that it before time t are blue. That in itself is did. Everyone concerned with induction problematic. or counterfactual reasoning has to take it Goodman argues that not all predi- seriously, not only nominalists. Recog- cates are projectible; so not all are suita- nizing natural kinds does not solve the ble for using in induction or projecting problem. Goodman’s challenge to realists from known to unknown cases. Predi- is: even if there are natural kinds, what cates like ‘green’ are projectible and justifies you in thinking that ‘green’, predicates like ‘grue’ are not. Although rather than ‘grue’ is one of them? all emeralds observed before t are grue, A further problem with Nominalism we cannot make inductive inferences and its Aftermath is that it mischaracter- from that fact to the conclusion that all izes Goodman’s epistemological stance. emeralds are grue. The critical question Shottenkirk states that Goodman «main- is what accounts for projectibility. tains a modified coherentism» (77). Al- Shottenkirk holds that Goodman’s though she admits that he is not strict answer is a form of social constructivism. coherentist, she holds that he simply al- «Predicates are entrenched because they tered a coherence account of knowledge. are in accordance with the practice of the This is not an accurate description. His- community» (85). This is incorrect for torically, epistemologies were character- several reasons. Goodman claims that a ized as foundationalist or coherentist. predicate is entrenched if it and its cog- Foundationalism maintains that knowl- nates have been projected far more often edge is grounded on beliefs whose cer- than any rivals it has. Non-projective tainty derives from experience or reason. uses of predicates are irrelevant. Because Coherentism maintains that knowledge he includes cognates, his analysis is not is grounded in a given belief’s function- dependent on any particular community. ing within a coherent system in which That ‘green’ is entrenched depends not each belief supports and is supported by only on the use of ‘green’ in projections the others in that system.