This article, perhaps more than others, is rather an exercise of linguistic logics. Roderick Chisholm, in “The Logical Form of Action Sentences, criticized an earlier article of Davidson, “The Logical Form of Action Sentences,” proposing a different theory. Davidson analyzes Chisholm’s theory in this article, trying to dismantle it.
Davidson discusses, in this article, the ontologies of events as unrepeatable particularities (“concrete individuals”). In this context, he recalls a previous work in which he proposed an analysis of propositions about particular events and actions. Roderick Chisholm commented on that approach by developing a new theory of ontologies in case of recurrence, claiming that it solves the problem more effectively. Davidson analyzes Chisholm’s theory, concluding that the theory of events as particularities proposed by Chisholm does not solve the problems highlighted in his own work: Chisholm gives us an idea of how interesting it would be if such a theory could be successfully developed.
Things are changing; but is there something like the change? Our language encourages us to believe that it exists, providing not only the proper singular terms, but also the whole apparatus of defined and indeterminate articles, tribal predicates, counting, quantification, and identity statements. If we take this grammar literally, if we accept these expressions and phrases as having the logical form that they seem to have, then we are devoted to an ontology of events as unrepeatable particularities (“concrete indivualities“). These are events of which at least one exists and is said to be identical to another. If dated, the particular events that appear to be necessary if such sentences are true stem from the principles of individual implicit (for example) in counting.
Events as particulars cannot, after all, be essential to our understanding of the world.
It would be better to judge whether we had a coherent and comprehensive account of the conditions in which our common beliefs (or those believed to be true) are true. If we develop such a theory and this theory would have called for a field of particular events, and we have not found any theory that works as well without events, then we have everything we can have to say that events exist; we have all the reasons we could say we say the events exist.
We do not have such a comprehensive theory, but we can learn by trying. A few years ago, Davidson proposed an analysis of the sentences about events and actions that involved a universe containing (inter alia) particular events. This analysis faces a variety of problems in what seems to be an attractive way.
Roderick Chisholm highlights a problem that Davidson has not discussed and, to solve it, he proposes a new theory, claiming that his theory solves the problem more effectively than Davidson.
Chisholms addresses “recurrence. . . the fact that there are some things that are repeated or happening several times.”
A natural way to provide an appropriate entity would be to say that one and the same event was instantiated every time. Or we could say that a member of the event class took place every time. Chisholm notes that these analyzes require two types of events, universals or classes, on the one hand, and instances or members on the other: he hopes for a larger ontological economy. His own solution is to invoke repeated events that can be said to happen, although there are strictly no such things as their occurrences. The repetition of the event would be treated in this way: the event took place both before and after this first occurrence. Davidson, talking about the ontological parsimony, wonders what’s so good about it? Clarity is desirable, but parsimony may or may not bring clarity. Of course, once there is a viable theory, it is interesting to say that only part of the ontology we thought was necessary would be sufficient; but this reduction must occur after providing a theory of work. It may also have insights to suggest what entities should appreciate in interpreting a certain part of the linguistic territory. To sum up classes or universals to explain recurrence, even if there are classes, Davidson would prefer a theory that did not invoke them here.
Davidson agrees with Chisholm’s goal of giving an account of recurrent events that do not require two distinctly different events. But he believes that special, unrepeatable events can help. One way is this: events have parts that are events and the parts can be temporally or spatially discontinued. So, the sum of all events is a special event, part of which took place at a time, and another part took place another time. We need three events to do this, but they have the same ontological status.
Even if only special, unrepeatable events are allowed, it is possible to give literal meaning to the statement that the same event occurs on two or more occasions. It is possible; but is this strange number of events really what we mean when we say, “the same thing happened again”? We can talk in some cases about the same event that continues after a pause. Davidson considers this to be forced, and the reason is that we normally do not need a single entity as a reference to resort to each use of the same thing. But the problem would be easier to solve if we identify “the same thing” with “something similar” or “another”. The recurrence in this case cannot be more than just similar, but distinct events, one after the other.
Thus, talking about the same recurring event no longer requires an event that happens several times. This shows at best that the theory of events as particulars is “”adequate to the fact of recurrence”.
To deduce some events from others we need premises. But then no analysis of these sentences in terms of events can be correct, which does not preserve the validity of inference. Davidson implement an example of this idea in logical inference of sentences, concluding that, when establishing random theory in response to the question, what kind of events, if any, must be if our statements must be true, we should be ready to test the theory by considering what implications it shows between assertions, or failing to show.
Davidson denies Chisholm’s argument: if it were a good argument, we could dismantle other true identity statements.
Since the events for Chisholm are entities that correspond to sentences or phrase-like structures, it is very apt to implement a proof that there are only two events in total – the one that occurs and the other. The problem is a standard one, to admit substitutivity of identity while denying extensionality. It can be done, except that Chisholm has not formulated a theory to do in case of statements about events and actions.
An appropriate theory should explain the adverbial changes.
It is not clear how Chisholm’s theory can handle the issue of the adverbial change. To take account of recurrence, Chisholm says there is only one event of every kind. To articulate the inference from one event to another, the start event must be complex. But it can not be the same as the resulting event, because the second could repeat itself, and the first one does not. It’s a contradiction.
About Chisholm’s attempt to analyze what appears to count, singularity and quantification, without the expected entities being counted, individualized, or quantified: At the center of the test are two concepts: the negation of an event and an event occurring before of the other (or the same). Both concepts require additional explanations. Since the denial of the event differs from the usual denial, we need an explanation of the non-p’s semantic role based on the meaning of ‘p’. Chisholm insists that the occurrence of non-p should not be confused with the absence of p. But this leads to contradictions.
The difficulty of “p occurring before the occurring of q” is that it does not suggest how to analyze this situation in terms “before” and “occurrence”. But until this is done, the inference from “p occurring before the occurrence of q” to “p occurs” cannot be proven to be valid. Given the individual events, the problem is easily solved: “There is an x event, an y event, and x occurred after y,” and this obviously implies the occurrence of each event.
Davidson thinks we must come to the conclusion that it is not clear whether a viable alternative to event theory as particulars can be developed according to the lines proposed by Chisholm. Difficulties depend, in most cases, on the fact that we have placed one interpretation or another on its suggestions; there may be other ways to develop the theory we missed. Chisholm’s approach deals with events that are just as mass terms, and many of the tribulations in analyzing specific sentences are similar to the problems that arise in trying to give a semantics of the terms of the table. There must be a correct way to do this in terms of mass, and if we know this, we could adapt this method to talk about events. Chisholm’s book gives us an idea of how interesting it would be if such a theory could be worked out successfully.
Source: Donald Davidson, Events and Particulars, Nous, Volume 4, Issue I (Feb., 1970), 25-32.