Away with words: commentary on the Atlan-Cohen essay 'Immune information, self-organization and meaning'

Drawing on metaphors from linguistics and information theory, Atlan and Cohen challenge us to take a very different view of the immune system, one that engages in constant chatter among the constituents and allows the immune system to arrive at a decision about what to, and not to, destroy. Our commentary responds to this challenge and points out many logical biological flaws in their view. We seem to agree that specificity is important, and that there is some kind of somatic selection process at work to distinguish self from non-self. Our analysis of models depends on the basis of how self and non-self are separated. There are only two possibilities, time or space; and space-based models are all but ruled out. There are two major kinds of time-based model, one based on the time taken for an organism to develop from embryo to adult, the other based on the time taken for a cell to differentiate from one state to another. With so many ambiguities in the metaphors and so little attention to mechanism, the Atlan and Cohen challenge is, we suspect, based on time measured in cell differentiation units. They also make the common mistake of assuming repertoires that are transcendental in size (>10(10)), making it impossible to have a functional immune system in animals smaller than a rabbit--a feature that does not instill confidence in the biological relevance of such models.