Analysis and Perspective
Origins & Design 17:1

On the Design of the Vertebrate Retina

George Ayoub
Department of Biology
Westmont College
Santa Barbara, CA 93108-1099

Abstract: It has been commonly claimed that the vertebrate eye is functionally suboptimal, because photoreceptors in the retina are oriented away from incoming light. However, there are excellent functional reasons for vertebrate photoreceptors to be oriented as they are. Photoreceptor structure and function is maintained by a critical tissue, the retinal pigment epithelium (RPE), which recycles photopigments, removes spent outer segments of the photoreceptors, provides an opaque layer to absorb excess light, and performs additional functions. These aspects of the structure and function of the vertebrate eye have been ignored in evolutionary arguments about suboptimality, yet they are essential for understanding how the eye works.

Editorial Introduction: A Popular Argument

It has been widely argued in both the technical (Thwaites 1984, Williams 1992) and popular evolutionary literature (Diamond 1985, Dawkins 1986, Miller 1994) that the vertebrate eye is poorly designed. "In fact it is stupidly designed," writes the influential neo-Darwinian theorist George Williams, "because it embodies many functionally arbitrary or maladaptive features" (1992:73). Chief among these features, Williams claims (1992:72), is the inversion of the retina.

"The retina is upside down," he writes. "The rods and cones are the bottom layer, and light reaches them only after passing through the nerves and blood vessels." These structures, claims UCLA evolutionary biologist Jared Diamond (1985:91)

aren't located behind the photoreceptors, where any sensible engineer would have put them, but out in front of them, where they screen some of the incoming light. A camera designer who committed such a blunder would be fired immediately.

The capstone of this argument is held to be the cephalopod (squid and octopus) retina, which is putatively "wired correctly," with its photoreceptors facing towards the light, and with its nerves "neatly tucked away behind the photoreceptor layer" (Miller 1994:30; see also Diamond 1985:91 and Williams 1992:74). The cephalopods, it is said, got it right.

In considering this argument, we may dispense immediately with optimality comparisons between cephalopod (invertebrate) and vertebrate retina designs. None of the authors cited above provides any evidence that the cephalopod retina is functionally superior to the vertebrate retina: a claim that, in any case, seems unreasonable on its face. Would hundreds of thousands of vertebrate species -- in a great variety of terrestrial, marine, and aerial environments -- really see better with a visual system used by a handful of exclusively marine vertebrates? In the absence of any rigorous comparative evidence, all claims that the cephalopod retina is functionally superior to the vertebrate retina remain entirely conjectural. In short, there is no reason to believe them.

But we should consider a more basic point. Why refer to the cephalopod retina at all? The claim that the cephalapods got it right assumes that the vertebrates did not, and that the latter are making the best of a bad situation -- but, of course, it remains to be demonstrated that, in fact, the vertebrate retina is suboptimal.

And this has not been demonstrated: not by the authors cited above, nor by other evolutionary biologists. "One of the difficulties with the hypothesis of optimality," note Farnsworth and Niklas (1995:355), "is the availability of observations to test it." That goes as well for hypotheses of suboptimality, as exemplified by the evolutionary literature on the vertebrate retina. The biological world is full of puzzling systems. While it is not readily apparent why vertebrate photoreceptors face away from the light, nor why other cell layers intervene, a good many things in science are not apparent at first glance. We need to look more deeply.

In this case, we need not look far. There are excellent functional reasons for vertebrate photoreceptors to be oriented as they are. These aspects of retinal structure and function have been ignored in evolutionary arguments about suboptimality, yet they are essential for understanding how the eye works.

-- The editors

The Structure of the Vertebrate Retina

First, some anatomy. Figure 1 depicts a vertebrate eye in cross section. Light passes first through the cornea, the primary focussing element, then through the iris, which controls how much light will enter the eye, and lastly, through the lens, which provides the adjustable focussing element. The light, now adjusted for intensity, is focused onto the thin tissue lining the back of the eye: the retina.

The retina (see Figure 2) comprises cells from the central nervous system (CNS), and converts or transduces light into electrical signals, the "medium" of the CNS. A highly complex tissue, the retina contains cells of several different types:

Photoreceptors (the rods and cones, labelled as "R" and "C" in Fig. 2) which actually convert the light energy into an electrical signal. These are the first cells directly involved in communicating information within the visual system, sending signals via a chemical synapse to:

The Bipolar Cells (B in figure), the second order cells in the retina; the bipolar cells synapse onto:

The Ganglion Cells (G in figure), at the inner surface of the retina; these cells have axons which travel together from the eye, exiting via the optic disk, and forming the optic nerve en route to the brain.

The Amacrine Cells (A in figure) mediate lateral interactions, transmitting information between adjacent bipolar cells and ganglion cells; and the horizontal cells (H in figure) which communicate laterally in the outer retina.

Now observe the path of light in Figure 2. The light must pass first through all the auxiliary cells before arriving at the photoreceptors -- which at first glance hardly seems sensible. If the design problem to be solved by any eye is forming a maximally accurate image of the world, then degrading the light before it reaches the "business end" of the photoreceptors seems self-evidently a poor solution. "This is equivalent to placing a thin diffusing screen directly over the film in your camera; it can only degrade the quality of the image" (Goldsmith 1990: 286).

And that is where evolutionary accounts leave the story.

The Critical Role of the Retinal Pigmented Epithelium (RPE)

But there is much more to be said. Lying directly behind the retina is an epithelial tissue which maintains the photoreceptors (see Fig. 2). This tissue, called the retinal pigmented epithelium (hereafter, RPE), is critical to the development and function of the retina. Indeed, volumes have been dedicated to understanding the role of the RPE (see, for instance, Steinberg 1985, Zinn and Marmor 1979), because when it malfunctions, the eye as a whole malfunctions.

1. Regenerating photoreceptive pigments

When light strikes a photoreceptor, it sets in motion a chain of molecular events which eventually culminate in forming an image in the brain. Here, let's focus on just the first characters in the story.

The first player on stage is the photosensitive molecule rhodopsin. Rhodopsin consists of a protein, opsin, and another molecule, 11-cis-retinal. Found at the distal ends of the photoreceptors -- the portion closest to the RPE, called the outer segment -- rhodopsin is embedded in membranous discs. [An important note about some potentially confusing terminology. The "business end" of the photoreceptor cell, where the membranous discs occur, is called the outer segment. In vertebrates, however, this segment is actually inner, i.e., at the back of the retina, pointing in towards the center of the organism.] When light strikes rhodopsin, the energy changes the shape of its molecular component 11-cis-retinal, into an all-trans conformation, a process called isomerization. This conformational change in retinal starts a complex cascade of reactions in several other molecules, causing the hyperpolarization (or shift in electrical charge) of the outer segment membrane. Molecular transmitters then carry this electrical signal from the synapse at the photoreceptor's base to the next neurons, the horizontal cells and bipolar cells -- thus beginning the process by which we see.

This process depends critically on the isomerization of 11- cis-retinal. Each photon of light striking a photoreceptor can isomerize retinal, and since many billions of photons constantly strike the eye, retinal must be replaced regularly to maintain the cycle, and overall photoreceptor function. That job of replacement falls to the RPE. The RPE cells collect the used retinal from the photoreceptors, and employ vitamin A to make fresh retinal, transporting it back to the photoreceptors (Bridges 1989; Hewitt and Adler 1994).

2. Recycling of photoreceptive material

Next on the list of RPE responsibilities is a related function: recycling the used outer segments. Outer segment membranes are very active, and thus must be continually replaced.

Each day, new outer segment membrane grows at the base of the outer segment (where it intersects with the inner segment, the cell region containing the nucleus), adding to the length of the photoreceptor. As the outer segment lengthens from its base, its distal end -- the oldest membrane -- sheds in segments. These segments are picked up by the RPE, which phagocytizes the material, recycling all of the molecules present (Bok and Young 1979).

Thus, spent photoreceptive membranes are removed from the optical path, to be replaced by new material. This process, which goes on continually, maintains the high sensitivity of the photoreceptors (Bok 1994).

3. Absorption of excess light

In addition to these active functions, the RPE also has an important passive role. Because it is heavily pigmented, it forms an opaque screen behind the optical path of the photoreceptors.

It thus absorbs light which is not collected by the photoreceptors, light which would otherwise decrease the resolution of images. This absorptive property of the RPE is important to maintaining high visual acuity.

This brief summary does not exhaust the functions of the RPE. Note, for instance, that the RPE "is required for the normal development of the eye" (Raymond and Jackson 1995), a function that, while not directly related to vision, certainly undergirds the very possibility of seeing at all. In short:

Considering the diverse functions of the RPE cells...there is no doubt that the integrity of the RPE metabolic machinery is essential for the normal functioning of the outer retina. Because of the nature of these interactions, it is essential that the RPE and photoreceptors be in close proximity for normal retinal function (Hewitt and Adler 1994: 67).

There are excellent reasons for vertebrate photoreceptors to be oriented as they are.

A Thought Experiment

But still: there sits a blind spot in each retina. To be sure, the blind spots are displaced laterally from each other, so that "with both eyes open, we can see everything in the visual field" (Williams 1992: 73), as one eye sees what the other does not. However, we can imagine situations where this wouldn't work:

Our retinal blind spots rarely cause any difficulty, but rarely is not the same as never. As I momentarily cover one eye to ward off an insect, an important event might be focused on the blind spot of the other (Williams 1992: 73).

So, as a thought experiment, let's fix the blind spot. We will start by turning the photoreceptors around, so their wiring isn't in the way.

We have eliminated the blind spot, providing slightly better sight in one portion of the eye. Now, however, the blood vessels and RPE, needed to maintain the photoreceptors, must be located on the inner side of the retina, between it and the lens. This places a large capillary bed (containing many red blood cells) and an epithelial tissue in the path of the light, significantly degrading the visual information passing to the photoreceptors.

Furthermore, since the photoreceptors continually shed material from their outer segments, dumping this opaque waste in the path of the light would greatly diminish the amount of light reaching the photoreceptors. Our proposed change also reduces the quality of the light, by refracting it with the opaque pieces of shed outer segment membrane.

We might imagine simply placing the RPE at the back of the retina, but this raises the problem of how to dispose of spent outer segment membranes, so that the photoreceptors can be quickly regenerated. Or, perhaps, we could surround each photoreceptor cell by RPE cells, but this would need increase the space between the photoreceptors, thus decreasing the resolution of vision.

These design changes may force temporal or spatial decrements in vision.

Are these improvements? Hardly; indeed, our thought experiment has taken the vertebrate eye rapidly downhill. In trying to eliminate the blind spot, we have generated a host of new and more severe functional problems to solve. Our "repair" seems far worse than the apparent flaw we wanted to fix.


The vertebrate retina provides an excellent example of functional -- though non-intuitive -- design. The design of the retina is responsible for its high acuity and sensitivity. It is simply untrue that the retina is demonstrably suboptimal, nor is it easy to conceive how it might be modified without significantly decreasing its function.


  1. Bok, D. 1994. Retinal photoreceptor disc shedding and pigment epithelium phagocytosis. In S.J. Ryan, ed., Retina, 2nd ed., Volume 1: Basic Science & Inherited Retinal Disease (St. Louis: Mosby), pp. 58-71.
  2. Bok, D. and Young, R.W. 1979. Phagocytic Properties of the Retinal Pigment Epithelium. In The Retinal Pigment Epithelium, eds. K.M. Zinn and M.F. Marmor (Cambridge, MA: Harvard University Press), pp. 148-174.
  3. Bridges, C.D.B. 1989. Distribution of retinal isomerase in vertebrate eyes and its emergence during retinal development. Vision Research 12: 1711-1717.
  4. Dawkins, R. 1986. The Blind Watchmaker. New York: W.W. Norton.
  5. Diamond, J. 1985. Voyage of the Overloaded Ark. Discover, June, pp. 82-92.
  6. Farnsworth, K.D. and Niklas, K.J. 1995. Theories of optimization, form and function in branching architecture of plants. Functional Ecology 9:355-363.
  7. Goldsmith, T.H. 1990. Optimization, Constraint, and History in the Evolution of Eyes. Quarterly Review of Biology 65: 281-322.
  8. Hewitt, A.T. and Adler, R. 1994. The retinal pigment epithelium and interphotoreceptor matrix: Structure and specialized functions. In S.J. Ryan, ed., Retina, 2nd ed., Volume 1: Basic Science & Inherited Retinal Disease (St. Louis: Mosby), pp. 58-71.
  9. Miller, R.F. 1994. The physiology and morphology of the vertebrate retina. In S.J. Ryan, ed., Retina, 2nd ed., Volume 1: Basic Science & Inherited Retinal Disease (St. Louis: Mosby), pp. 58-71.
  10. Raymond, S.M. and Jackson, I.J. 1995. The retinal pigment epithelium is required for the maintenance of the mouse neural retina. Current Biology 5: 1286-1295.
  11. Steinberg, R.H. 1985. Interactions between the retinal pigment epithelium and the neural retina. Documenta Opthalmologia 60:327-346.
  12. Thwaites, W. 1992. Design: Can we see the hand of evolution in the things it has wrought? In Evolutionists Confront Creationists, Proceedings of the 63rd Annual Meeting of the Pacific Division, AAAS, Volume 1, part 3 (San Francisco: Pacific Division, AAAS), pp. 206-213.
  13. Williams, G.C. 1992. Natural Selection: Domains, Levels, and Challenges. Oxford: Oxford University Press.
  14. Zinn, K.M., and Marmor, M.F. 1979. The Retinal Pigment Epithelium. Cambridge, MA: Harvard University Press.

Copyright © 1996 George Ayoub. All rights reserved. International copyright secured.
File Date: 6.22.96