Brief of Work in the area of and

• Basic Questions (from Squire and Kandel) • how does memory work • are there different kinds of memory • what is their logic • where in the brain do we learn • where do we store what is learned as memory • can memory be resolved at the level of individual nerve cells

Prior to the 19 century

The early philosophical approaches used methods of , logic, and argument to: • define broad constraints on knowledge and what was knowable • determine what was innate or acquired

The late 19th century brought quantitative study of mental processes

• The 1885, Hermann Ebbinghaus monograph On Memory described systematic measurement of memory in terms of accuracy, retention, and capacity. This work also revealed some basic properties of memory such as differences in duration and the impact of repetition on retention. • In 1887 , Sergei Korsakoff published his work on a syndrome that established the use of memory disorders as a means of studying processes.

1 • In 1890, wrote Principles of that outlined sharp distinctions between short (primary) and long (secondary) term memory. • 1898, Edward Thorndike working under James published his work using animal subjects to study memory.

The 20th century brought the behaviorists demonstrating systematic ways of modifying behavior. • Around 1904 Pavlov published his work on , or the conditioned reflex establishing the basis for systematic control of behavior through learning. • In 1911, Thorndike demonstrated the principles of trial-and-error learning or instrumental/operant conditioning. • The work of John Watson, B.F. Skinner, and Clark Hull in the 1930’s and 40’s introduced highly systematic singular black-box rules of learning for explaining complex behavior.

A more cognitive approach evolved from the limitations apparent in the behaviorists paradigm

• Tolman (1949) argued for a more cognitive mechanism of learning in which different types of learning used different rules. He studied more complex tasks that required behavioral selection related to problem solving. • In the 1950’s Frederic Bartlett examined memory of naturalistic scenes and stories in an effort to examine the cognitive complexities of memory.

2 In addition to this work examining the principles of learning at the behavioral level there was an effort to identify the mechanisms responsible for learning and memory at the neural level. • In the 1890’s following his pioneering work in brain anatomy Ramon Cajal suggested that structural changes in the brain might underlie learning and memory. • In 1897 Charles Sherrington named the "synapse" and suggested that it was a likely candidate for a structure related to learning and memory.

• Beginning in the 1920’s Karl Lashley began studying the effect of removal of different amounts of cerebral cortex on maze learning. From this work the idea of "mass action" was formulated in which brain function was non- localized. • In 1938, the neurosurgeon Wilder Penfield demonstrated that electrical stimulation of the brain, in particular the temporal lobe, could elicit reports of , , and hallucinations including voices, images, and music.

• In 1949, Donald Hebb published his work the Organization of Behavior in which he argued that memory was held in cell assemblies distributed throughout the brain which was in part, a response to the earlier work of Lashley. • In 1957, Scoville and Milner published their work examining the patient H.M. in which removal of regions of the medial temporal cortex led to profound loss in memory capabilies without loss in intellectual or cognitive abilies demonstrating clear functional heterogeneity within the brain in contrast to the earlier notions of Lashley.

3 • Subsequent work in the 1960’s by demonstrated that in H.M. was unaffected further demonstrating that different types of memory could be maintained by different brain regions. • In 1967, Seymour Benzer working in the fruit fly Drosophila using chemical mutagenesis to alter the function of individual genes implicated specific proteins in behavior and introduced the use of molecular genetic approaches in the study of memory.

What is Learning and what is Memory?

A Definition • Learning is an adaptive change in behavior resulting from experience.

From this, the early behaviorists black-box view can be seen as

Experience ------‡ behavior 1 ------‡ behavior 2 ------‡ behavior 3

INPUT BLACKBOX OUTPUT in which selection is based on consequence (e.g. reinforcement, Skinner).

•In non-associative learning, the adaptive advantage is already established. •In more associative learning the extent of adaptive advantage must be evaluated.

4 Further refinement of this model in a reductionist direction can be described as

Experience -‡ neural activity -‡ neural output pattern 1 -‡ behavior 1 -‡ neural output pattern 2 -‡ behavior 2 -‡ neural output pattern 3 -‡ behavior 3

INPUT BLACKBOX OUTPUT

In which selection of neural patterns is based on consequence (e.g. reinforcement, Hebb)

Let us examine a simple definition of memory A Definition •Memory is [a process related to] the retention of learning. •A function of memory is to allow production of learned/adaptive behavior at a later time. •But the adaptive response may not be the same as the original learned behavior (when does learning occur?). •Note that our simple description of the learning process involved the selection of behavior based on the evaluation of consequence. •Memory also involves a process by which past experience can be used to guide future (novel) adaptive behavior.

Introduction to behavioral learning paradigms: • Instrumental conditioning • avoidance learning - receive an unpleasant stimulus when the animal fails to make a response - 2-sided shock chamber • learned helplessness - inescapable shock results in freezing - used in context dependent fear conditioning • reward training - receive reinforcement when response is made - shaping • escape - inevitable unpleasant stimulus continued unless response is made • punishment - receive an unpleasant stimulus when a response is made

5 Classical conditioning • can be subdivided into appetitive conditioning when the unconditioned stimulus is rewarding, or defensive conditioning if the stimulus is aversive. • Examples of classical conditioning paradigms are eyeblink/nictitating membrane taste aversion GSR

Serial learning • lists • sequences • mazes • Paired associates - positive and negative stimuli.

Spatial learning • radial maze

Biological constraints Conditioning • stimulus/response preferences - some stimuli are more easily associated with some responses • easy to jump or bar press for food; hard to condition to grooming or scratching • shock produced fear easily paired with visual/auditory stimuli; difficult to pair with taste • illness easily associated with taste but not visual/auditory stimuli • avoidance of shock easily paired with barrier jumping but not bar pressing.

6 • rats are good at radial maze but pigeons are poor, yet pigeons are good in open field tasks. • foraging patterns in rodents takes relative food quantity into consideration • instinctive drift observed in highly trained animals

Ethology • - study of behavior in relation to the environment; innate behavior neuroethology - the study of the neural basis of innate behavior • - echolocation in bats • - electroreception in fish • - sound localization in owls

Factors that influence learning

• Repetition -> predictability • Magnitude -> salience • Order/timing -> causality • Reinforcement -> significance/valence

Can these factors be related to mechanisms at a cellular level?

7