Chapter 2



The Lesion Experiment

Subtractive Logic in Pharmacological Experiments


Anatomical Destruction

Biochemical Disruption


Development of Brain Structure

Emergence of Behavior and Brain Chemistry




Return to Main Table of Contents



The Lesion Experiment

The rationale of traditional physiological psychology is based largely on the fact of neuroanatomy. The brains of closely related species have similar appearances, even when viewed in some detail. The brains of different individuals within the same species bear an even more striking similarity-- a fact that has been termed neuronal specificity:

"The location of nerve cells, the trajectory of nerve fibers and the spatial array of synaptic connections are invariant in all individuals of the same species....In this way, the main neuronal circuits achieve an architecture that is breathtaking in its complexity, but frugal in its variability."

(Jacobson & Hunt, 1965, p. 26)

The obvious presence of systematic organization within the brain encouraged the early students of behavior to seek specific relationships between structure and function. The roots of physiological psychology can be traced rather specifically to a series of 19th century experiments by Flourens (1824). By observing the effects of surgical damage, Flourens established the general functions of the spinal cord, the cerebellum, the medulla and the hemispheres. At about the same time, the more detailed physiological studies of Bell and Magendie (ca. 1825) revealed the anatomical and functional distinctions between sensory and motor nerves. Somewhat later, Fritsch and Hitzig (1870) used electrical stimulation techniques to show the important principle of topographical organization, the point-to-point mapping of brain cells and the muscle groups that they serve. The results of these studies were readily accepted, not so much because they were without contradiction, but more because of the Zeitgeist for classification schemes and molecular analysis (cf., the cell theory and developments in chemistry).

All of these experiments were conducted within the framework of the subtractive model of structure/function relationships. The principal method of testing this model is the ablation experiment, which attempts to determine missing behavioral capacities following the surgical removal of specific regions of the nervous system (i.e., the "subtraction" of a portion of the brain to determine which aspect of behavior is missing).

The common sense appeal of the structure function notion probably would have been sufficient to sustain the use of the ablation experiment, but a combination of technological advances and key experimental findings pushed its popularity almost to the point of uncritical acceptance. In 1908, Horsley and Clarke invented the stereotaxic instrument which made it possible to direct surgical damage to specific structures deep within the brain. In 1939, this instrument was modified for use on the rat (Clarke, 1939), setting the stage for a series of impressive demonstrations of the ablation experiment. Hetherington and Ranson (1940) showed that specific destruction of the ventromedial nucleus of the hypothalamus reproduced the so-called clinical obesity syndrome which, previously, had been associated only very loosely with tumors of the hypothalamus or pituitary gland. Soon it was shown (e.g., Teitelbaum & Stellar, 1954) that specific destruction of the neighboring lateral hypothalamus (less than a millimeter away) resulted not in obesity, but in the rejection of food to the point of starvation.

Thus, the stage was set for at least a quarter century of attempts to relate specific areas of the brain to specific behaviors. During recent years, this method has fallen into disrepute as methods that appear to be more sophisticated have been developed. The problems are not restricted to the technical deficiencies of the lesion method, and we may as well face them now before we go into some of the pharmacological data.

One of the most frequent criticisms of the subtractive logic method is the tendency to ascribe functions to "holes" in the brain. Technically speaking, a lesion in the auditory cortex does not cause a loss of hearing, it is that the remaining portions of the brain cannot mediate the function of hearing. This point of logic seems to be at once subtle and obvious, but the fact remains that errors in interpretation of these experiments are quite likely. The common example of the naysayers is a hypothetical experiment which would first show that a frog will hop when a loud noise is made. This behavior would disappear after removal of the legs. The conclusion: Frogs hear with their legs. The real point to be made here is that this example seems absurd only because we have so much independent knowledge about frog legs and hearing. We would not have this luxury if we attempted to discover, for example, the role of the stria terminalis in eidetic imagery.

There are two major reasons why we should not cavalierly abandon the subtractive model of structure/function relationships and the ablation method of testing this model. The first reason is the simple assertion that function must have a structural basis, and that this structural basis lies within the brain. Specific functions may not be closely tied to obvious structural organization, but in the final analysis, some set of physical structures will be held accountable for behavior.

The second reason for holding onto the subtractive model of determining structure/function relationships is, again, simple and compelling: subtractive logic is the only method available. Flourens did not discover the subtractive model and the ablation method, but rather applied the only tools and logic that the scientific community is willing to use to the specific problem:

(a) an assumption that the universe is lawfully organized (brain structure is related to brain function),

(b) a belief that proximal cause and effect relationships exist (brain structure actually accounts for brain function), and

(c) that cause and effect relationships can be determined by manipulation (changes in structure produce changes in functions).

Subtractive Logic in Pharmacological Experiments

It is important to realize that pharmacological experiments are a direct extension of the ablation method and subtractive logic. The logic being applied is that if a certain brain transmitter is blocked and certain behaviors change, there is a (chemical) structure/function relationship (refer to Fig. 2-1). All of the more sophisticated chemical interventions (e.g., the mimickers, enzyme blockers, alpha blockers, and others as outlined in chapter 3) are simple variants of this. Some will argue that stimulation experiments (as opposed to lesion or chemical blockade) circumvent these logical problems, but the arguments are less than compelling. A brief review of figure 1-10 will demonstrate the complexities that are involved. Stimulation of the parasympathetic nerve, for example, results in either stimulation or inhibition of the muscle, depending on where one measures the response. Likewise, chemical stimulation with NE results in either stimulation or inhibition of the muscle, depending on where one measures the response. Stimulation experiments are neither simple nor straightforward in their logic, and must be viewed with the same eye for complexity as other experimental approaches.

Even developmental and comparative studies can be characterized by what might be referred to as a receding subtractive model. The logic is based on the fact that young animals that do not have fully developed behavior potential show a common emergence of structure and function, including the emergence of specific pharmacological systems. If time went backward, it would be a lesion study. In fact, the lesion study does occur in nature as time goes forward, with changing behavior as a result of changing structures during senescence. The ablation method, subtractive logic, and the associated problems are pervasive. The effects of alpha-adrenergic blockade, precommissural fornix transection, or the development of olfactory capabilities are no more or no less "messy" than those of Lashley's (1929) classical experiments that attempted to specify the brain mechanisms of intelligence.

The point of all of this discussion is to encourage you to approach the field of behavioral pharmacology with the same squinty-eyed skepticism that has been applied to the old-fashioned lesion methods. It is a relatively new field of inquiry, but only in terms of the procedures. Highly sophisticated approaches and details of neurochemistry do not alter the simple fact that the interpretation of these results is subject to the same pitfalls as the interpretation of brain lesion experiments. And a healthy level of skepticism will lead us to more detailed and accurate accounts of the relationships between drugs and behavior. Let us examine a set of clinical and experimental results that encompass a wide range of anatomical, pharmacological and developmental observations to see how these methods may be applied.


Anatomical Destruction

One of the most bizarre cases of accidental brain injury is that of Phinnaeus P. Gage, a railroad worker (See Bloom et al 1985 for a detailed account). In the Fall of 1848, Gage and his crew were blasting rock. The procedure involved drilling a hole in the rock, then stuffing the hole with alternate layers of packing material and black powder. The packing material was tamped into place with a long steel rod. In a moment of carelessness, Gage apparently tried to tamp the powder layer, and a spark ignited the powder. The resulting explosion transformed the tamping rod into a four-foot projectile which entered Gage's left cheek, passed through the top of his head, and landed several feet away (Fig. 2.2).

To the amazement of Gage's crew, he sat up and began talking to them minutes later and, after a short trip by ox cart ambulance, was able to walk up the stairs to his hotel room. The attending physician determined that his entire index finger could be inserted into Gage's brain case. He placed bandages on the wounds, and followed him through an uneventful period of healing.

Within a few weeks, Gage's physical recovery allowed him to return to his job as foreman. He had no apparent intellectual deficits or memory losses. Yet, his return to work quickly showed the nature of the deficit that follows massive frontal lobe damage. The formerly mild mannered, thoughtful and cooperative foreman had been transformed into a cursing, belligerent tyrant. He lost his job, joined a traveling sideshow for a few years to capitalize (in a small way) on his misfortune, and died of an epilepsy attack about 13 years later.

Phinnaeus P. Gage's tragic case history has become famous in the annals of abnormal psychology as an example of altered behavior resulting from traumatic brain injury. Thousands of other case histories involving trauma, cerebrovascular accidents, tumors and the like have provided parallels to animal experimental data to build up a fairly accurate picture of the functional relationships between anatomy and behavior. Although the effects of such damage are never as simple as one would like, there is nonetheless a certain degree of predictability of motor disturbances, sensory problems, intellectual deficits, emotional disturbances, memory losses, and so forth.

Biochemical Disruption

In the same way that Gage's case history serves as a useful parallel to experimental lesion studies, the disease known as phenylketonuria can serve as a useful parallel to experimental pharmacology (Kaufman, 1975). Phenylketonuria, usually referred to as PKU, is a genetic disorder that appears to follow the simple laws of Mendelian genetics. It is estimated that about 1 in 50 persons is a carrier of the recessive gene for this disorder. When two such carriers have children, there is a 1 in 4 chance that the disorder will be manifested (on the average, 2 in 4 will be carriers; 1 in 4 will be normal). As a result, the incidence of the disorder is about 1 in 10,000 births.

Phenylketonuria appears to result from a disorder of a single gene which controls a single enzyme which transforms the amino acid phenylalanine into another amino acid, tyrosine. When this liver enzyme is missing or seriously deficient, phenylalanine accumulates in the blood, and related metabolites called phenylketones begin to appear in the urine (hence, phenylketonuria). In the absence of treatment, the prognosis is very bleak indeed. The patients typically show severe mental retardation with an IQ less than 30, abnormalities of posture and reflexes, reduced pigmentation, and a life span that rarely goes beyond 30 years.

The details of this disorder, which are still unfolding, can provide some valuable lessons in the study of brain chemistry. In particular, it serves as a model of the complexity that must be considered in any pharmacological manipulation. The first and perhaps most obvious manifestations of PKU can be attributed to the effects of neurotoxins. In the absence of the critical enzyme, phenylalanine levels rise to about 15 times their normal level and allow large quantities of phenylketones to be produced. These substances may have some direct toxicity to brain cell physiology, but more importantly, they appear to interfere with chemical transmitter systems (Fig. 2.3). A series of different neurochemicals called catecholamines (including norepinephrine, DOPA and dopamine) share tyrosine as a precursor. In the absence of the enzyme, there is less conversion of phenylalanine into tyrosine and deficits in the production of these transmitters may occur. Another way of looking at this is that the PKU deficit makes tyrosine an essential amino acid, since it can only be obtained from dietary sources.

Ironically, the current evidence suggests that this direct effect may be less important in the resulting mental retardation than another, indirect effect. Another essential amino acid, tryptophan, is converted into serotonin (5-HT), an important neurotransmitter of the central nervous system. There is some degree of overlap in the biochemical pathways that convert tyrosine and tryptophan into their respective transmitters. As a result, high concentrations of phenylalanine may inhibit the formation of serotonin. In untreated cases of PKU, serotonin levels are dramatically low. The most successful treatment of PKU is the dietary restriction of phenylalanine which prevents the build up of phenylalanine and related metabolites. This dietary restriction produces a rapid elevation of serotonin levels and halts the progression of the disorder. If such treatment is begun at birth, the major signs of retardation are averted, although some minor difficulties always seem to remain, presumably because of prenatal factors.

The impaired function that results from untreated PKU cases is completely analogous to the case of Phinnaeus P. Gage--it is the result of localized brain damage. In the case of Gage's traumatic lesion, the damage can be localized by an anatomical description. In the case of PKU patients, the damage is localized by a biochemical description, i.e., a biochemical lesion of the serotonin system and, probably, the catecholamines. The important point to remember is that drug effects and transmitter disorders are no more or no less than physical lesions or electrical stimulation.


Development of Brain Structure

The development of the human brain is nothing less than awe inspiring. Between the time of conception and the time of birth, the full complement of some 100 billion individual neurons are formed, migrating to their appropriate place within the brain structure. This number is perhaps a little easier to comprehend when one considers that, on the average, a million new brain cells are formed each four minutes between conception and birth! At the time of birth, all of the cells have been formed, but the development of synaptic connections continues, perhaps for 30 or more years in humans. Each neuron communicates with something on the order of 1,000 other neurons and may have anywhere from 1,000 to 10,000 synaptic connections on its surface (103 synapses times 1011 cells equals 1014 synapses with an even more staggering number of possible synaptic interactions.)

The orchestration of all of these connections during the process of development is further complicated by the presence of numerous different chemical transmitter systems. Obviously, our meager abilities to analyze the nervous system can hardly scratch the surface of this process, but we can, nonetheless, gain some general insights into the important events that accompany the development, maturation and ultimate degeneration of the brain.

In the first section of this chapter, we mentioned in passing that the developing brain could be compared to a lesion experiment. Prior to the complete development of the brain, any behavior that occurs must be mediated by the part of the brain that has developed, without the part that is to be developed--exactly as in the case of Phinnaeus P. Gage, whose behavior depended on the remaining brain tissue. If the development of the brain were completely random, the study of developing behavior would provide little information. But to the extent that specific systems may develop in sequence, we may gain some information by looking for parallels between the onset of certain behavioral abilities and the appearance of certain brain systems.

It is important to keep in mind the types of changes that we must look for in the developing brain. Even in the case of human brains, which are relatively immature at the time of birth, all of the cells are present and, to a large extent, in position. This does not mean that the brain is functionally mature at birth. In some cases, the maturation of the biochemical machinery that is necessary for chemical neurotransmission is not completed until some time after birth. Furthermore, the physical complexity of some cells, especially the formation of synapses, may continue for many years after birth. Thus, there is an interdigitation of both anatomical and chemical development.

In general, the increasing anatomical sophistication of the developing brain follows a plan that has been termed encephalization (see Fig. 2-4). The spinal cord and brain stem regions develop and mature at an early age, while midbrain and forebrain regions develop later. The cortex, and particularly the frontal regions, are the last to show fully developed anatomical complexity. These anatomical changes cannot be ignored, but for the present purposes, we will concentrate on the development of neurochemical systems.

Emergence of Behavior and Brain Chemistry

One of the recurrent themes in developmental psychology is that of the emergence of behavior. The term emergence has been used advisedly, because it brings forth the image of something popping to the surface of a pool, rather than the piecemeal, brick by brick construction of some edifice. Although a case can be made for both types of appearances of behavior, the more remarkable case is that in which the relevant behavior or set of behaviors suddenly appears in full-blown form. A common example of this is the motor skill of walking in human infants. No amount of "training" can teach an infant to walk at age six months, and a complete lack of training does not appreciably retard the age of onset of walking. The "skill" usually emerges within a week or two following the first tentative steps.

We are currently at a very primitive level of understanding the emergence of more complicated, cognitive behaviors. (Unfortunately, the accumulation of this knowledge appears to be more of the ilk of brick by brick construction, rather than some epiphany popping to the surface.) One of the most complete stories that we have available at this point is that involving the neurotransmitter acetylcholine. There are many gaps in the data, but if we are willing to fill these with a bit of conjecture, we can see a coherent example of the parallel emergence of behavior and neurochemical function.

The story began with a theoretical review by Carlton (1963), setting forth the notion that behavioral excitation was under the control of brain systems that use norepinephrine as a neurotransmitter, whereas behavioral inhibition was under the control of brain systems that function through acetylcholine. Accordingly, an increase in activity could result either (a) directly, as a result of stimulation of the norepinephrine system, or (b) indirectly, as a result of blocking the acetylcholinergic system. (We will learn more about the specific drugs later, but for the present purposes a general description will suffice.)

Campbell and associates (1969) extended our understanding of these theoretical notions by studying the development of these behaviors in young rat pups. When rat pups are about 14 days of age, their ears have been opened for only a few days, their eyes are just opening, their fur just barely conceals their skin, and they are very cute. Even at this early age, a significant increase in their activity is produced by drugs that stimulate the catecholamine (norepinephrine and related compounds) systems. By contrast, attempts to increase activity indirectly by blocking acetylcholine functions was to no avail. By 18 to 21 days of age, manipulation of the cholinergic system began to have an effect, and by 28 days of age, the cholinergic blocking effect increased activity in the same way that it does in adults.

The interpretation of these experiments suggests that the brain systems that use acetylcholine as a neurotransmitter are not yet functional in the 14-day old rat. Obviously, drugs that interfere with such a system could not be effective, because there would be no substrate for them to work upon. Once the system has become functional at three to four weeks of age, the drugs have a substrate to work upon and become effective.

It should be noted here that the results of these experiments using a combination of drugs, behavior and age point toward a very specific and testable conclusion: Namely, that a detailed look at the brain should show the maturation of acetylcholinergic cells when rats reach the age of three to four weeks. The results of neurochemical assays (Matthews et al, 1974) provided confirming evidence for the late development of the acetylcholine pathways. Furthermore, related experiments that involved both behavioral and neurochemical assays showed the emergence of brain systems that utilize serotonin as a transmitter at about 10-12 days of age. Thus, we begin to see a general tendency for the sequential development of systems that can be defined on the basis of the neurotransmitter.

The behavioral functions that are mediated by emerging cholinergic systems appear to be far more complicated than the locomotor activity that was measured in these early experiments. As Carlton suggested in his early review (1963), the cholinergic systems seem to mediate a more general inhibition of behaviors that are non-rewarded or actually punished. Behavioral studies in numerous laboratories, including that of the authors (cf., Hamilton & Timmons, 1979), have buttressed this view, showing that the ability to perform tasks that involve behavioral inhibition do not reach maturity in the rat until about three to four weeks--the same time that the neurochemical systems emerge.

The neurochemical systems must be located somewhere, and this location defines the anatomical basis for these same behaviors. Although the neurochemical systems are somewhat dispersed in the brain, the acetylcholine fibers are heavily concentrated within the limbic system and the forebrain (cf., Feigley & Hamilton, 1971). Literally hundreds of experiments have shown that physical damage to these areas impair the ability of animals to inhibit behaviors that are punished or non-rewarded (cf., McCleary, 1966).

In the case of humans, these inhibitory abilities become very sophisticated and involve such behaviors as inhibiting the attainment of immediate gratification because of complex social contingencies that are likely to be in effect in the future (e.g., not eating the chocolate cake that has been prepared for tomorrow night's birthday party). Interestingly, the inability to meet such exigencies is characteristic of "childish" behavior, and inhibitory control emerges in rudimentary forms at about 4 to 6 years of age. From that point on, the development of these abilities seems to be more gradual, perhaps continuing until about 30 years of age. It is perhaps more than coincidence that the anatomical complexity of the forebrain region shows a concomitant development, and that many of these fibers utilize acetylcholine as the neurotransmitter.

Let us return for a moment to the case of Phinnaeus P. Gage. Despite massive physical trauma to the forebrain, he did not exhibit straightforward deficits in intelligence. Nor did the literally thousands of patients who underwent frontal lobotomies during the unfortunate period of history when these were in vogue--IQ scores typically remained the same or actually increased by a few points (cf., Valenstein, 1973 for related discussion). Thus, man's proudest possession, the big frontal lobes, seem to be minimally involved in intelligence scores. But, they are involved in foresight and the social control of behavior that will have future consequences. This may be the unique feature of humanity, and it may involve to a large extent, the neurons in the limbic system and frontal lobes that function via acetylcholine.

If we follow this line of argument and the associated experimental evidence through adulthood into senescence, the story continues to hold together. One of the emerging characteristics of senility is that of so-called "childish" behavior. A closer examination of these behaviors reveals that it is the failure to take into consideration the future effects of one's behavior upon others--a failure to respond to social contingencies that require the inhibition of behaviors. It has been known for some time that the aging brain shows selective deterioration of small fibers of the cortex. Evidence from autopsies performed at Johns Hopkins (e.g., Coyle et al, 1983) have shown that the majority of the fibers that deteriorate seem to be those that utilize acetylcholine as the neurotransmitter, especially in the case of Alzheimer's disease.

Thus, the small cholinergic fibers that are the last to appear are also the first to be lost (see Fig. 2-5). The behavioral capacity for complex inhibition is the last to appear and the first to be lost. In the middle, interference with acetylcholine neurochemistry or destruction of the areas where these fibers are concentrated results in impairment of inhibitory behaviors. All of this may be more than coincidental.

We have seen in this chapter some general approaches to the analysis of brain and behavior. We go now to some of the details of the methods that are involved, and will return later to show the application of these details to particular interfaces of behavior and brain chemistry.



1. All brain research methods, including drug studies, are modeled after the subtractive logic of the lesion experiment.

2. Certain metabolic disorders may interfere specifically with individual neurotransmitter systems.

3. The major neurotransmitter systems mature at differing rates, with the behaviors that are controlled by these systems emerging as cells become functional.

4. Senescence may involve the specific decline of some neurotransmitter systems while others remain more or less intact.


Ablation experiment


Alzheimer's disease

Behavioral inhibition





Neuronal specificity





Subtractive model