Jablonka & Lamb (J&L) address the puzzling question of how the same genotype can yield different phenotypes in different environments (Jablonka & Lamb Reference Jablonka and Lamb2005). An extension of this problem is how a genotype can create a well-adapted organism in a variety of environments, and how this can be accomplished in the human with only about 30,000 genes. One of the secrets is that the process of development uses more information from the environment than anyone had envisioned. It is a kind of one-generation Lamarckism: the environment tuning the phenotype in adaptive directions, using genetic systems that assemble information from the environment rather than providing the information themselves. Rather than an inheritance of acquired characteristics, it is acquiring characteristics from the environment, guided by inheritance. Parts of organisms must be competent to use environmental information, though, even if the information is not in the genes.

A relatively simple example of environmental influence on phenotypic development is in the vertebrate eye. The genes could have guided the development and growth of a precise geometry for the eye, forming a focused image just at the retinal surface. This would require a lot of genes working in just the right way, and the last bit of precision would require most of the genetic instruction. Instead, the eye is formed in a general way by the genes, and is made too small. Another genetic mechanism gives the command, “grow until focus is good, then stop.” Thus, the eye achieves precise focus without precise instructions (Wallman et al. Reference Wallman, Turkel and Trachtman1978).

Another use of environmental information became clear when I had the privilege to participate in the discovery of visual cortex receptive fields (RFs) that are far outside the normal range, due to rearing in an abnormal visual environment. Helmut Hirsch of the Stanford University psychology department had raised kittens through the critical period for RF development wearing masks that presented a pattern of large vertical stripes to one eye and large horizontal stripes to the other, focused at optical infinity. Before Hirsch performed behavioral testing for his Ph.D. dissertation, Nico Spinelli of the Stanford Medical School psychiatry department suggested that our team in Professor Karl Pribram's laboratory record RFs from the primary visual cortex of the kittens, using extracellular microelectrodes. Spinelli, Robert Phelps, and I had been recording from single neurons of cat and monkey cortex with an automated method for more than a year (Spinelli et al. Reference Spinelli, Pribram and Bridgeman1970).

For our first recordings, Hirsch brought a kitten to our laboratory in a light-tight box (the kittens had been dark-reared except for a few hours of mask exposure per day). Our technique scanned a 25×25 degree field with a small moving spot in a raster scan. The first RFs were diffuse and ill-defined. That evening, a RF appeared on our screen unlike any we had ever seen – it was more than 20 degrees long, and monocular. We checked the stimulation apparatus and the recording, and found everything normal. The RF orientation happened to be in the direction of the scan, so a spontaneous burst of firing might have caused the result. We then rotated the scan direction 90 degrees and mapped the field again, but it appeared just as before. We looked at each other, dumbfounded: this kitten's cortex had a completely different organization!

Later that night, we recorded several more such RFs, all monocular, all huge, and all following the orientation of the kitten's mask. Other kittens yielded similar results (Hirsch & Spinelli Reference Hirsch and Spinelli1970). The implication is that the cortex had reorganized itself to produce RFs reflecting the structure of the visual world the cortex encountered. It was the first indication that the environment could not only bias the statistics of existing RFs, but could also create completely new ones never seen in nature.

J&L provide a context for examining the implications of this and subsequent results. If a bizarre environment can induce bizarre RFs, what induces normal RFs? The only answer consistent with the fields found in the mask cats is that the environment is doing the same thing in the normal case as in the mask case. The statistics of the environment become reflected in the structure of the visual system's RFs, which therefore become adapted to best code events during the animal's lifetime. According to this idea, the RFs in the cat cortex, and by extension in monkey and human as well, are the way they are because they are tuned by the environment during early development.

The properties of the normal visual world, then, are reflected in the structure of RFs in normal cortex. The world contains contours at all orientations, but more in the horizontal and vertical directions, similar to the RF distribution. In the spatial frequency domain, 1/f power describes both the world and the RF distribution. Motion is ubiquitous in the environment and in RF sensitivities. Like the optics of the eye, the structure of cortical connections is very general, requiring very little genetic information. The high-resolution information comes from the world, not the genes. The system is adapted to adapt.

The arrangement can be described as one-generation Lamarckism because acquired characteristics, from the structural regularities of the environment, come to define what had previously looked like fixated, genotype-directed development. Unlike the original Lamarck proposal, the environmental information does not make it into the next generation, but the genetic strategies supporting the competence to acquire these characteristics do.

We do not know how far one-generation Lamarckism extends. What would be the color sensitivities, for example, of cortical RFs in animals raised in monochromatic or bichromatic light? We know that we are stuck with genetically specified receptor pigments, but the cortex may be a different story. There's lots of work to do, using what Hirsch calls “environmental surgery” to investigate the extent and importance of environmental influences on phenotype development.