Talk:Fallacy of composition

From Wikipedia, the free encyclopedia

Socrates
Philosophy Portal
This article is within the scope of the WikiProject Philosophy, which collaborates on articles related to philosophy. To participate, you can edit this article or visit the project page for more details.
Stub This article has been rated as Stub-Class on the quality scale.
Mid This article has been rated as mid-importance on the importance scale.

I enjoyed the example of saving = good for individual, saving for everyone = bad for the whole. This seems to be a prisoner's dilemma if someone wants to perhaps add it in (although it only works for examples where there is choice among people, it doesn't work for the machine breaking example). --ShaunMacPherson 19:21, 12 July 2005 (UTC)

I think this principle is incorrectly stated, because in some (many) cases properties of component parts are retained by the whole. For example, metal melts, therefore if a machine is constructed of metal parts the machine will also melt. It depends on the specific properties and how the parts are combined; in fact there is no general method for determining whether the fallacy applies or not, you need to understand the details of the case in question. Bobcousins 23:51, 22 May 2006 (UTC)

Bobcousins, the fallacy is correctly explained in the article. Your example shows only that some inferences from the properties of parts to the properties of wholes happen to issue in true conclusions. But so, for example, do many hasty generalisations issue in true conclusions. The point in the case of fallacy of composition is that more evidence is required than simply the evidence that the parts have certain properties. Try this argument, in which premise C provides such extra evidence:
A. Ice must always melt above 0ºC.
B. This statue is made only of ice.
C. Melting is a property that transfers from parts to wholes.
Therefore D. This statue must melt above 0ºC.
This argument is valid. But the argument A, B, therefore D is not strictly valid, even though the conclusion D is plainly true. This reduced argument is an example of the fallacy of composition. Noetica 00:26, 23 May 2006 (UTC)
And perhaps another way to put it is that the Fallacy of Composition is indeed a "fallacy" because although the conclusion may sometimes be true, it is not necessarily true in every case. 24.6.66.193 12:36, 9 July 2006 (UTC)
Ok, I appreciate your response. But I still disagree it is correctly explained. The conditional "when" is attached to "inference from part to a whole", but that is not where the fallacy arises. Without further qualification, this implies that all such inferences are fallacious, but this is not correct. It's a first step, but the fallacy arises "when that property is not transferable". The principle of transferability is a fundamental condition, therefore the definition should read: "A fallacy of composition arises when one infers that something is true of the whole from the fact that it is true of some (or even every) part of the whole AND that property is not transferable". Otherwise the definition amounts to "a fallacy arising by assuming something is true when it is false" which is correct, but not explanatory, and leaving the reader to deduce the principle of transferability from the example.
For a better text see http://www.nizkor.org/features/fallacies/composition.html, which includes the essential point about characteristics
Bobcousins 11:57, 14 November 2007 (UTC)
I assume you mean the "when" in this clear introductory statement, Bob, since it is the only occurrence of that word in the article: A fallacy of composition arises when one infers that something is true of the whole from the fact that it is true of some (or even every) part of the whole. You say Without further qualification, this implies that all such inferences are fallacious, but this is not correct. On the contrary, though: it is correct! If my support for my conclusion that object O has property P is just that some (or all) parts of O have P, I have fallen victim to the fallacy of composition. It doesn't matter if O does have P; it doesn't even matter if O has P in virtue of its parts having P: the fallacy is still operating. If, on the other hand, my evidence is that the parts of O have P and that P is transferable from parts to wholes, that's a different story. It is not the story given in our first, definitional, sentence: ...from the fact that it is true of some (or even every) part... . Contrary to what you say above, the fallacy does not arise just "when that property is not transferable": it arises when such transferability is not among the evidence presented in the premises of our argument.
The source you cite says something similar to this, in fact, right up front: ...when, in fact, no justification [is] provided for the inference. That's right: the justification (concerning transferability) may be potentially available, but if it is not there in the evidence presented in the premises, you get the fallacy of composition. Further on, your source says: In some cases, sufficient justification can be provided to warrant the conclusion. I think your source is careless or frankly mistaken in failing to clarify that the sufficient justification must in fact be provided, if the fallacy is to be avoided. It is not enough that it can be provided, in some imagined alternative way in which the argument might have been fashioned.
Specifically, I dispute this example (along with much else, not relevant to the present discussion): [I]t is true that an individual rich person has more wealth than an individual poor person. In some nations (such as the US) it is true that the class of wealthy people has more wealth as a whole than does the class of poor people. In this case, the evidence used would warrant the inference and the fallacy of composition would not be committed. Not so. The conclusion is only warranted when you explicitly add the demographic and statistical evidence referred to. Truth of the conclusion, along with partial support from the premises, does not confer validity; nor does it confer freedom from fallacy.
– Noetica♬♩ Talk 13:49, 14 November 2007 (UTC)
(Nu?) In fairness though, Bob, I have taken one side in a rather complex debate. The view of the highly respected logician and Irving Copi is like mine, but others support something like your view. This article (along with its sibling, Fallacy of division) is important enough to expand so that it reflects the range of opinions. The whole matter of classifying fallacies – and articulating precisely how they operate, and how they are to be distinguished from other fallacies – is mired in difficulties. When I have the energy and time I may do some systematic work on the articles myself. Not yet, though!
– Noetica♬♩ Talk 08:43, 18 November 2007 (UTC)

I don't like the economic example. There are many economists who do dispute this Keynesian's idea. Actually what you see is that if one person saves, he decides not to consume. His decision somehow translates to the economy in the sense that some businessman doesn't sell. If many people save, many people cease to consume, the decision of many people affects the economy and many businessmen fail to sell. I believe this is Keyneses fallacy of treating small effect as a non-existent effect. Second, from the economic point of view, if the businessmen did expect that the consumers would want to save, they would have altered their investment decisions and they wouldn't have produced the goods they can't sell and there would have been no recession. What acutally creates the recession is that the producers didn't expect such behaviour. The whole idea that if many people save that it is bad for economy is in itself an economic fallacy, I believe Henry Hazlitt did some work to expose these Keynesian fallacies. 85.207.163.148 (talk) 23:16, 16 February 2008 (UTC) Ondrej Palkovsky

[edit] Economic

I am not familiar with the "paradox of thrift". But as described, it sounds to me like this is an exmple of prisoner's dilema. What is the relationship between falacy of composition and the prisoner's dilema?