Meeting 2.2: Life Without Definitions
Cleland, Carol E. (2012) Life without definitions. Synthese 185:125-144
Summary: Lucas Mix presented a summary of the article.
Thesis: “a scientifically compelling understanding of the nature of life presupposes an empirically adequate scientific theory (vs. definition) of life” Abstract
“Traditional” Definitions provide necessary and sufficient conditions for using a term
Natural scientists want a definition to agree with a natural kind.
Analysis of how terms are used will not bring us closer to finding a natural kind. (!)
Theoretical Definitions include “empirically revisable, provisional theoretical claims” (126)
Current scientific theories are unreliable.
Alchemists defined water by its ability to act as a solvent. “But as we now know the alchemists were wrong.” (!) Water is H2O and that knowledge requires a framework: molecular theory.
We have reason to believe that alien life would be truly alien, but we don’t know how alien. Thus we have no equivalent theory of life. We currently have a sample size of 1.
Stipulative Definitions “assign new concepts to old or new terms” arbitrarily (131)
Water = H2O is really a ThD, but it is the poster child for SD
Descriptive Theory (attr. John Locke): “concepts are identified with descriptions qua logical conjunctions of predicates supplying necessary and sufficient conditions for the application of the term.” (131-2)
But then, we cannot claim a definition is wrong, only that the terms have changed.
Newtonian “mass” is a special case of Einsteinian “mass”. This means Newton was not wrong (but definitions can be?).
Putnam’s arguments against DT: On twin Earth XYZ has the same sensible properties as H2O on Earth, but a different molecular formula. Observers on the two planets use the same term, but they are talking about different things. Therefore, the extension of the term water is not wholly in the mind.
“A theory of meaning should be neutral with respect to what our scientific theories tell us about the world.” (?)
Causal Theory of Reference: “the extension of a natural kind term is fixed not by human concepts but the actual nature of the things the term designates.” (135) But how do you orient yourself?
Cluster Theory: (Wittgenstein) attributes a term to a cluster of related predicates. Fails essentialist test.
Causal-Descriptive Theories: assign terms to phenomena that frequently co-occur
“The fact that biologists cannot confidently constrain the possibilities for life strongly suggests that our present-day concept of life does not contain identifying knowledge of the kind. (138)
Water = H2O turns out not to be a theoretical identity statement, because the terms are not the same. (138)
[Doesn’t this contradict the idea that the definition of water is not SD on page 131?]
We can bring concepts into closer alignment with the world.
Corundum includes both rubies and sapphires, while Jade is really nephrite and jadeite.
[But that doesn’t deny the original categories were natural…]
There is currently a preference for microstructural essences.
In astrobiology, we should be looking for anomalies. “we need to search for physical systems that both resemble familiar life in striking ways and also deviate from it in provocative ways.” (141) [Which ways?]
We need “tentative criteria.”
Critique by Lucas
“Without additional examples of life, one cannot discriminate features that are universal to life, wherever it might be found, from features deriving from mere physical and chemical contingencies on the early Earth…” How do we know that there is other life to compare it with?
Isn’t there an epistemological distinction that can be made between stipulative definitions and the natural kinds they attempt to model? Thus 17th century ‘water’ may be logically valid (and even useful), without being as useful as the 20th century ‘water.’
In regards to Putnam’s arguments for essentialism: If two things have the same sensible properties, are they not empirically the same thing?
Aren’t provisional definitions necessary for communication? Should we stop saying “life”?
What work does a definition of life do?
This was the main take home message from the conversation. What ends does a “definition” serve and how do we judge if they are served by a particular definition. Astrobiology has a clear operational objective: find life elsewhere. Other uses seem less clear. Cleland makes a compelling point about current definitions getting in the way of the astrobiology objective.
Another issue arises in scientific communication and reproduction of experiments. It is often trivial to say that a component of an experiment is alive rather than dead; nonetheless, the distinction is often important to the outcome. What is the significance of this qualifier that makes it useful in the trivial cases?
Outside of science, is value attached to the definition of life?
Does the category of “life” represent an important area of research?
Life appears to be uniquely associated with the process of natural selection and adaptation. Research into these areas has been fruitful. To what extent should the two be identified? Life also appears to be associated with certain types of complexity and structural organization. To what extent should life be identified with organizational properties?
What is the best way for scientists to look at definitions?
Objectivity arose again and the question of whether empiricism excludes “internal observable states”.
How do cluster definitions (a la Wittgenstein) relate to definitions that treat life as a syndrome?
Is a syndrome definition useful?
In medicine, a syndrome is a disorder characterized by a number of observable symptoms rather than a causal mechanism.
Concern was mentioned about whether either the syndrome definition would be helpful if the symptoms were poorly defined.