Dear colleagues and friends:
Loreta and I devoted recently some thought to field homology in a wide conceptual context (Brain Res.Bull.57:243-255 /2002). I don't know whether Harvey or Erich have had opportunity to read that, but if you give a glance to it you will see how we envisage a conciliation of both lines of interest in the task of defining homologies today.
In essence, we agree with Harvey that it is fundamental to push forward as far as possible, seeking homologies even at the cellular level (.....I also fully agree with him that we may spend fruitlessly our time if we haggle too much on terminological niceties -homology is the issue). However, we think that the cellular approach needs to stand upon a firm ground of previous field homology, and therefore the latter approach should not be presented as an alternative or opposed option, at least in the way we conceive it. Please don't let Harvey's passing comment of fields as "ill-defined" convince you that you are unable to define a better field yourself. History of neuromorphology certainly includes ill-defined fields, as well as it does ill-defined anatomical boundaries (i.e., limits of the hypothalamus), ill-defined sulci and nuclei, ill-defined cortical layers (i.e., Wulst "layers"), and, why not, ill-defined cell types (i.e., spongioblasts). Can we believe that any past or present errors signify no success was ever reaped at any level? Past failures need to be taken in our stride, so to speak, learning from them if we can. Moreover, the essence of the field concept is that it does not need a visible limit to exist as a causal entity (how well-defined is the Earth's gravitational field? What about a neuroepithelial field defined by a particular combination of genes expressed, irrespective of where its cellular derivatives stabilize after migration, or its diffusible secreted molecules do? -note this uses cellular properties).
I'd like Harvey to be more explicit, so that we can understand why he seems to believe he can define good homologies at the cellular level, without previous consideration of the field homology of the brain region where the cells are, or come to be. You cannot evade considering a topological position somewhere in the brain wall of the cells you are comparing (both before and after any migrations). I believe exclusion of attention to this important aspect possibly entails either strong belief in previous field formulations (as unexplored assumptions), or, less probably, conscious/subconscious eschewing of undesired interpretative possibilities (we all have our pet theories there, at risk of being wrong). On the other hand, anyone progressing to a sound field homology postulate and confessing momentary ignorance of more detailed cell homology possibilities, does not stop others from advancing in that direction. I know Harvey's thought must be deeper than implying that restricted consideration of cell typology amply defined (but without any topology) is enough by itself to ascertain a sound homology relationship. Perhaps if he gives us an example of what he considers proper cell-level homology, irrespective of what field homology may have to say, then we can discuss the point more precisely. Even if we do not agree on names, finally, having these discussions is fun!
Take care, Luis