Psych 322: Biological Basis of Learning and Memory

Maine de Biran

(1804) *Envisioned 3 types of memory: 1. Representative memory -everyday "memory" -what we will call declarative memory (kind of like facts from our day or how we remember the day) 2. Mechanical memory -unconscious memory for actions -we now know involves several types of memory (probably meant reflexes and habit memory) 3. Sensitive memory -unconscious emotional memory -we now know to be supported in part by the amygdala *he was also a philosopher, not a scientist *did not consider the biology of how memory worked (just conceptually) *one of the first people to talk about the different types of memory and not just memory in general

Franz Gall

(1825) -originator of phrenology (the detailed study of the shape and size of the cranium as a supposed indication of character and mental abilities) -thought different parts of brain supported different functions -thought that one could discern another's strong traits by feeling bumps on the skull -imagined that some brain areas were important specifically for memory -this is the only technology that Gall had available to him even though it may seem dumb and primitive now -Gall thought that memory for words and the capacity for speech were very close to each other but two distinct regions in the brain (thought were in the front of the brain, prefrontal cortex) *thought you could determine if someone was good with words if their eyeballs were bulging out

Charles Darwin

(1859) -developed idea of natural selection as a mechanism of evolution -idea has an enormous impact on study of biology -idea that randomness and natural selection are at the core of life led scientific thought away from creationism -indirectly led to scientific investigation of memory -darwin did not come up with the idea of evolution, came up with natural and sexual selection as mechanisms of natural selection -did not directly lead to the modern day study of modern day memory but shaped all scientific thought across the world at this time -- went from naturalism to experimentation

William James

(1890) -philosopher who is often thought of as one of the first American psychologists -used others' investigations of reflex pathways as a starting point for thinking about memory -distinguished habits from true memory -distinguished working memory from long-term memory (he called these primary and secondary memory, respectively) -representative of transition from introspection to experimentation -good representative transition man of the change in modern neuroscience/understanding of the brain because both a philosopher and experimenter (psychologist) -starting writing about reflexes as a type of memory -know you needed data and experiments to have a traction about a topic

Karl Lashley

(1929) -asked whether memory for maze was stored in a particular brain region -trained rats to run mazes -carefully removed different parts of the brain in rats -retested brain damaged rats (after removing part of brain) to see if they could still remember the maze *found that particular location of damage DID NOT seem to matter *rather, found that HOW MUCH of brain was damaged was what mattered *concluded that all areas of cortex contribute to memory equally, that is, that memories were distributed -memories are placed in one place in the brain and everywhere at the same time *"angel falls in located in Venezuela" example -mice showed it did not matter what region of cortex he cut out *showed it mattered how much, not where *correlation between how much cortex was removed and the number of errors the mice made **the more damage or brain removed, the more errors the mice made *concluded memory did not lay in one particular place

B.F. Skinner

(1938) -one of the most prominent figures in behavioral psychology or behaviorism -argued that all behaviors could be explained by a series of stimuli and responses -unobservable mental events were best ignored -purposeful avoidance of brain and cognition set tone of psychology during middle of 1900s *focus on behavior became the predominant way to study memory -thus, influence of Hebb's ideas was not as great as one might have suspected they should be -believe you should focus on behaviors to look at memory -thought that "cell assembly" was too abstract and should focus on concrete observable behaviors

Wilder Penfield

(1938) -neurosurgeon who stimulated brains of awake patients during surgery to treat epilepsy (Penfield and "burning toast" clip) -stimulation in some areas elicited specific memories -finding that stimulating specific areas led to specific memories suggested that memories were LOCALIZED -Broca's area: important for speech production *interruption of speech when touched this area of the brain with electricity/probe -found that stimulating the brain could either impair or elicit particular memories so this suggested that memories do belong in one area

Donald Hebb

(1949) -suggested a reconciliation between Lashley's and Penfield's findings -suggested that memories are supported by networks of interconnected neurons *he called this type of network a CELL ASSEMBLY -"cells that fire together wire together" -talked about how there is local aspects to memory and distributed aspects, so BOTH

***Brenda Milner

(1957) -neuropsychologist who studied patient HM -HM had medial temporal lobes removed bilaterally as last effort to get rid of his epilepsy -surgery led to memory impairment that is called anterograde amnesia -however surgery left intact language, IQ, personality, perception, and working memory -HM's case indicated that medial temporal lobes were important for memory -HM's case was beginning of modern study of memory systems -Suzanne Corkin continued the study of HM

Terje Lomo and Timothy Bliss

(1973) -discovered long-term potentiation (LTP) -LTP refers to the strengthening of the connection between two neurons -thought to reflect a cellular basis of memory -LTP is one mechanism by which one group of neurons connection with another group of neurons can be strengthened

Baddeley's model of working memory

(1980s) -1980s version of what memory should look like -sub-vocal rehearsal refers to phonological loop -visuospatial sketchpad is the ability to use info in your mind without using sight *ex. close your eyes and try to remember all words on the board -the central executive in this model is like the agent and decider of how you are going to use the visuospatial sketchpad and phonological loops (decider of what you remember when your eyes are closed, etc.)

Aristotle

(ca 350 bc) -one of the earliest thinkers to contemplate the workings of memory -considered psychological distinction between recollection and familiarity *this distinction between recollection and familiarity plays an important role in memory research -was a philosopher, NOT a scientist -thought that the heart was the source of intellect -he was not trying to do experiments, he was just thinking logically how memory worked in his own mind/heart

Neuroimaging of 6 patients with brain damage vs. 2 healthy controls

(see camera roll for photo) Show here are magnetic resonance images for six of the seven patients and two healthy individuals. Patient AB was unable to participate in the MRI because he wears a pacemaker. However, he was included in the experiments on the basis of the neuropsychological profile indicating well-circumscribed amnesia and a high resolution CAT scan that was consistent with a finding of damage restricted primarily to the hippocampal region. These images are coronal images taken at about the level of the mammilary bodies. The hippocampal region is here on the right and on the left. It may difficult to see, but compared to controls, the patients all have visible damage to the hippocampal region. In contrast, the adjacent entorhinal, perirhinal, and parahippocampal cortices appear normal. To quantify these impressions, we measured for all six patients the volume of the hippocampal region and the adjacent cortical structure in the medial temporal lobe. For patient JS, the hippocampal region was not reduced in size, but several focal lesions were unmistakable. For the other patients, the average reduction in volume of the hippocampal region was 30% relative to controls. In contrast to the reduction of the hippocampal region, the average volume of the adjacent cortical structures was the same for patients and controls. The experiments I will present today included these seven patients and healthy individuals who were matched to the patients with respect to age, education, and the information and vocabulary subscales of the WAIS-III intelligence test.

Rey-O figures with the 7 brain damaged individuals and 2 controls from the brian scans

(see camera roll) -The right part of this figure shows the reproduction from memory by the seven patients of an abstract line figure after they had copied the figure 15 minutes previously. -Here is a drawing from memory by a healthy individual. -Below is the delayed reproduction by a patient whose brain was inspected by histological analysis and was found to have damage restricted primarily to the CA1 field of the hippocampus proper. Although the drawing is poor compared to the control's, RB was able to capture a modest amount of the figure. -Below that is the delayed reproduction by a profoundly amnesic patient with virtually complete damage to the medial temporal lobe. When asked to redraw the figure he had copied 15 minutes prior, EP asked "What figure?" and declined to venture a guess. -What one can see is that the delayed reproductions by the seven patients in the present experiments are more similar to the drawing by the patient with damage to the CA1 field than to the patient with widespread damage in the medial temporal lobe. All the patients remembered having drawn a figure previously and remembered at least that the figure was an abstract line drawing. Thus, the patients have a modestly severe memory impairment consistent with damage limited to the hippocampal region.

Episodic Memory

(type of declarative memory) -memory of events *single episode *personally-experienced (more emotionally-charged?) *Includes contextual details *ex. I ate a hamburger for lunch

Semantic Memory

(type of declarative memory) -memory for facts *knowledge about the world -not necessarily tied to original learning incident (not necessarily personal) *ex. a hamburger is made of beef -there is a distributed cognitive network of semantic memory, talked about in categories or subsets of topics (ex. animal -- fish -- salmon, etc.) *all over different regions of the brain

Graph that is showing news events recall vs. when the patient sustained memory damage

*LOOK AT CAMERA ROLL* -for both anterograde and temporally-graded retrograde amnesia, the orange box (normal memory, 15-30 years BEFORE damage to hippocampus) was very similar in brain damaged patients and healthy controls -impaired memory after brain damage and impaired memory 5-10 years before damage were both significantly worse than in the control patients (impaired memory after damage even more so)

Slide on perirhinal, parahippocampal, and entrohinal cortex fucntionings image

*check camera roll to study this diagram/ pattern

Cushing's syndrome

-Chronic stress hormones (e.g, Cushing's Syndrome) can negatively impact hippocampus

***Conscious vs. non-conscious memory & declarative vs. non-declarative

-Conscious = you bring to mind something -Non-conscious = you get better at something without necessarily bringing it to mind -Declarative memory = a form of long-term memory that involves the MTL; conscious memory -Non-declarative memory = long-term memory that is not declarative (hence the name "non-declarative memory); involving somewhere other than the MTL; non conscious memory

Damage to MTL results in:

-Anterograde amnesia -Temporally-graded retrograde amnesia *consolidation can explain temporally-graded retrograde amnesia -Intact working memory -Intact non-declarative memory -Intact non-memory abilities -Amnesia spares non-declarative memory (shown in the mirror-word task)

***Main points about amnesia lecture

-Damage to MTL results in profoundly severe anterograde amnesia (HM and EP) -Damage restricted to hippocampus results in moderately severe anterograde amnesia -Damage to MTL (or just hippocampus) leads to temporally-graded retrograde amnesia -Damage to MTL leaves intact working memory -Damage to MTL leaves intact non-memory abilities -Damage to MTL leaves intact examples of nondeclarative memory -One distinction between declarative and nondeclarative memory is distinction between conscious and nonconscious memory

Consolidation and retrograde amnesia

-Damage to MTL results in temporally-graded retrograde amnesia -The process by which memories that were once dependent on the hippocampus (and rest of MTL) become independent of the hippocampus (and rest of MTL) is called memory consolidation -Memory consolidation provides an explanation for the observation of temporally-graded retrograde amnesia

Professor's view on episodic and semantic memory

-Declarative memory depends on the hippocampus. -All declarative memories are acquired as part of event. -Initially, those memories are distinct and rich in detail for the contextual details. You could call them episodic details if you really wanted to. -Over time, those memories are consolidated into our existing fund of knowledge. -The aspects of the memories that made them distinct-the episodic details—are sacrificed so as to emphasize the aspects of the memories that are shared with other memories—the relationships between semantic features. -This process is useful. -Thus, I would say that the process of consolidation is also a process of "semanticization". -My main point: although the hippocampus seems ideally situated for acquiring episodic memory, a closer look suggests that the process of acquiring semantic memory also depends on the hippocampus.

**Emotion is not always beneficial to memory

-Defenbacher analyzed 21 studies on emotional accuracy, and found mixed results. However, he concluded from the bulk of the data that an optimal amount of stress is needed in order to ensure memory accuracy. -In addition, as can be seen from the diagram, simple stimuli seemed to be remembered easier, despite the level of arousal over complex or semi-complex stimuli. -This seems to establish that mood or emotion in general (see Eysenck or Niedenthal, Setterlund, and Jones (1994) for review; emotions: specific events with distinct expressions, moods: diffuse feeling, non-specific, and longer lasting. No specific expressions) has a strong effect on memory. What about emotional disorders?

EP video and past memory

-EP could easily recite directions to places in his childhood neighborhood but could not even tell you where he lives right now (he lives in San Diego) -this was published in a study -had a map of Hayward when EP was growing up in the 50's and had other people who grew up their and moved away take this test and see if they could navigate through the Hayward from their childhood -EP rembered his old neighborhood fairly well, did just as well on this memory task as the healthy patients *but cannot tell you anything about current hometown or geography *meaning his cognitive ability to navigate is perfectly in tact, but is memory is impaired

HM and his hippocampus

-HM's lesion includes cortical areas surrounding the hippocampus, not only was his hippocampus removed but so were surrounding parts of his MTL -perirhinal cortex, entorhinal cortex, and parahippocampal cortex (LOOK AT CAMERA ROLL)

Hippocampal and neutral activity

-Hippocampal activity correlates with memory for neutral material (and, not shown, emotional material) -A set of the same authors set out to further differentiate the roles of the amygdala and other MTL areas, specifically the HC in memory for emotionally-neutral words -8 males heard neutral words spoken during a PET session -Free recall of the words was conducted 24 hr later Correlations were conducted between rGMR and memory -Results showed that only activity in the left HC was associated with memory performance -Also, there was a lesser extent of activity in the right HC -There was no association between activity in either amygdala -This shows that the amygdala is not active during the encoding of neutral materials, but the HC is

Morris water maze

-In neuroscience, the Morris water maze is a behavioral procedure designed to test spatial memory. It was developed by neuroscientist Richard G. Morris in 1984, and is commonly used today to explore the role of the hippocampus in the formation of said spatial memories. -a rat in put in a pool of water and learns to be able to locate the hidden platform in the middle of the pool -the rat's successful escape reinforces its desire to live and thus is able to find the hidden platform the more trials it does -water maze experiment with rats with hippocampal damage *rats with hippocampal damage learned slowly but still eventually learned -healthy rats were able to express their memory for the goal flexibly (probe tests) while hipppocampal damaged rats were not

***Fear conditioning (with rats primarily) research more in-depth

-Philliips & LeDoux showed that contextual FC is mediated by a different pathway than tone-cued fear conditioning -They trained animals on the standard version of the task, then assessed freezing both immediately after the rats were placed in the conditioning chamber and in response to the tone -Amygdala lesions blocked conditioned freezing to both the context and the tone -By contrast, damage to the HC selectively blocked contextual fear conditioing, sparing the CR to the tone -So, if we combine these results with the brain structures previously discussed, the full set of circuits mediating fear conditioning in this task involves a set of parallel and serial pathways to the amygdala -Most direct is from auditory thalamus, secondary through thalamocortical, but contextual fear conditioning involves a yet more indirect pathway by which multimodal info arrives in the HC and is sent to the amygdala via the HC

Transitive inference

-Transitive inference (TI) is a form of deductive reasoning that allows one to derive a relation between items that have not been explicitly compared before -In a general form, TI is the ability to deduce that if Item B is related to Item C and Item C is related to Item D, then Item B must be related to Item D -healthy rats were able to express (flexibly) their memory for relations (TI), which rats with damage to the MTL did sig. worse in -rats with damage to MTL were able to order pairs of things however, just had trouble with relation piece -probe tests: hippocampal cognitive map encodes locations AND the points of interest at those locations -supports the idea that declarative memory is relational and flexible

Amygdala and hippocampus activity for positive emotional material

-amygdala (and hippocampus) activity correlates with memory for material that arouses positive emotion too -can get emotionally activated amygdala by both negative and positive things (but less likely to have positively charged emotional memories) -this study was showing gruesome images (severed legs, blood, etc.) associated with negative emotions activating the amygdala versus porn/sex associated with positive emotions activating the amygdala - equally activated amygdala but difference between imagery, negative vs. positive arousal

Copying down an (linear) image from memory -- Rey-O test

-basic neuropsych test = asked patient to make a copy of the drawings on this sheet of paper -reasons for this: *make sure you don't have a motor problem *when drawing this, it will help you form a memory of this -after 15 minutes, EP could not even remember that he drew a picture on a paper 15 minutes ago *EP declined drawing the picture another time because could not even remember drawing a picture -another patient similar to EP (similar brain damage), did not remember he drew a picture 15 minutes ago but said he would draw a picture anyways and drew a horse (because he likes horses) -are these patients aware of their memory impairments? *patients like HM and EP are not even aware that they have severe memory impairments like this (are not aware that their major brain damage is impacting their memory)

Consolidation

-consolidation of memory refers to the process that explains the phenomenology of temporally-graded retrograde amnesia -it is usually considered to consist of two specific processes, synaptic consolidation (which occurs within the first few hours after learning or encoding) and system consolidation (where hippocampus-dependent memories become independent of the hippocampus over a period of weeks to years) -a lot info that you hold onto for 5 or so seconds and than do not hold onto it *the process where our brains convert short-term memories into long-term ones -the neural pathway of how you bring old memories to mind (memories from 15 to 40 years ago) does not involve the hippocampus or the MTL *this is the process of how EP remembered how to get around Hayward from when he was a child, even after sustaining damage to hippocampus and MTL *this could be linked to the process of consolidation

Results from functional brain imaging in humans (fMRI)

-different areas of the brain important for different types of memory -encoding and retrieval from declarative memory via connections with MTL ***LOOK AT PHOTO TO STUDY THIS TOPIC***

Episodic and Semantic Memory

-episodic and semantic memory are two types of declarative memory 1. Episodic memory = memory for events - single episode - personally-experienced - includes contextual details - e.g., "I ate Philly rolls for lunch." 2. Semantic memory = memory for facts - knowledge about the world - not necessarily tied to original learning incident - e.g., "Philly rolls are raw salmon + cream cheese." -these two types of memory are separable but interrelated, innately connected -professor thinks episodic memory is very overrated -as humans, use episodic memories to make social bonds, accuracy is not always as important for us -hippocampus is important for both of these memories because both are declarative memory

How memory for emotional events is heightened due to emotional arousal (and release of hormones) during thta event

-idea that lots of areas able to make new associations with the amygdala - memories coming in and making essential plasticity in the amygdala (connecting the memory to fear, arousal, etc./ signal of a threat or a social cue, etc.) *yelling at us for texting in class example -how do you leverage these arousal cues to prioritize processing in other areas of your brain?

Studies in rats show the importance of amygdala activation for emotional memory

-if the rat goes into the cage and gets shocked when they go into the cage, they are less likely to want to enter to cage the next trial around *low norepinephrine = worse memory and more likely to go back into the cage on next trials *inhibitory avoidance increases amygdala norepinephrine (NE) levels for at least 2 hours *the level of NE in the amygdala correlates with the strength of the memory as measured by reluctance to enter shocked side of box -norepinephrine is an indicator of much the amygdala is modulating these other brain structures (specifically the hippocampus) *so more of this, means more modulation of hippocampus and better memory

Overactive amygdala in PTSD patients

-imaged PTSD patients, healthy controls, and people who went to war but did not come back with PTSD -in this machine, played random noises and then played combat sounds (machine, guns, etc.) --PTSD patients' amygdala was not saying "oh crap" for everything but when seeing action movie video of guns, their arousal went way up (should not have extreme arousal to seeing guns in a movie, but they did because this is where their PTSD originally stems from--combat)

flashbulb memories

-is a highly detailed, exceptionally vivid 'snapshot' of the moment and circumstances in which a piece of surprising and consequential (or emotionally arousing) news was heard *ex. people's memory of 911 or OJ Simpson trial -flashbulb memories may actually be more about confidence in your memories than actual accuracy in your memories -study was done about OJ Simpson trial and asked people specific questions like where they were than they heard the news, etc. *asked the same people these same questions 15 months later and 32 months later and recorded their answers *after 32 months, the majority of the sample had majority memory errors/distortions in their story (from original story they gave)

Why do we have the process of system's consolidation?

-it could require more resources for all memory to be stored in one place -so if one part of your brain gets damaged you don't lose all of your memory -need to free up some space in certain parts of your brain -kind of like the "cloud" - store it in the cloud in case your hardware crashes

Kaleidoscope recognition test example

-kaleidoscope recognition test = look at an image of a kaleidoscope and wait a certain amount of seconds and try to pick out this image from a bunch of other images of kaleidoscopes *you need to make a long-term memory of these because you cannot make a working memory of these images easily (making a declarative memory of it) *little ability to rely on sub vocal rehearsal and working memory in this experiment *how long memory could last determines long-term or working memory?

Remembering letters example in class..... (testing our short-term sensory store)

-letters from example: H Q O M K R *I got 6 out of 12 *in labs the mean is normally 4 out of 12 -experiment testing your short-term visual sensory store (that is what is accessing those letters in that moment) -you probably have access to all of the letters in your short-term sensory store but as you begin to write them down, about a third of the letters exit your short-term sensory store (what you actually have available to you) -fleeting after image when you close your eyes - the info out in the world is still accessible to you for a few milliseconds (short-term sensory store) -partial report (sound coming on the screen to signal one row of numbers) is not to get you to only look at the bottom row, but it is trying to tell your brain what information to drag out of your short-term sensory store and into your working memory -7 + or - 2 "rule" (how many letters one can typically remember, or lists of things -- that is why phone numbers are 7 numbers, minus the area code)

Hippocampus volume and memory

-lower hippocampal volume correlates to worse memory -reduction in hippocampal volume correlates with reduced memory

***Hebb's Cell Assembly

-memories depend on the connections between different points, or nodes, in the network -so is memory localized or distributed? BOTH. -to the extent that a memory depends on each individual connection, it is localized -but to the extent that recalling a full memory depends on reactivating the entire network of neurons, it is distributed -this concept is FUNDAMENTAL to the entire course, UNDERSTAND IT!! -this topic is kind of abstract so confusing for scientists to figure out -neuroscience was the dominant field at this time so took a while to figure out

Early philosophers

-no empirical data, largely introspective

Patient RB

-patient RB has damaged restricted to his hippocampus -patient RB had a case of moderately severe amnesia -used same drawing test where drew a picture and then x amount of minutes later asked the patient to draw the same picture from memory *RB remembered that he drew a picture and remembered enough to draw basic lines and shapes *concluding that memory impairment that results from just hippcampal damage is much less severe than damage to entire MTL **HM and EP could not even remember that they copied down an image at all

Selective damage to the amygdala eliminates the improvement in memory for emotional material

-pt. SM has selective damage to her amygdala, on both sides *this is a rare disease and probably something she was born with and increased as she aged *"HM patient of the amygdala world" *more more rare to have selective amygdala damage than it is to have selective hippocampus damage

Temporally-graded retrograde amnesia

-refers to a phenomenon of premorbid memory loss whereby information acquired recently is more impaired than information acquired more remotely

Retrograde amnesia and impairment of memories

-retrograde amnesia (cannot remember old memories) may extend back many decades in patients with large MTL damage -how far back in time can you grab info that you once know even if you have had damage (info that you stored decades ago may still be damaged) -info learned close to damage is the most limited and info learned farthest back from damage is the less restricted and the more likely you are to remember it

Behavioral psychology

-set back the field of memory science for a while because Skinner suggested that unobserved mental effects should be ignored

What is the difference between short-term and working memory?

-short term does NOT really have to do with how long you access that memory but it has to do with how long that memory could last (like how long it could stay a memory) *legit means how long a memory is STORED *ex. you heard something five seconds ago could be in long-term memory already -- onomonopea and plink example -short-term memory: not a durable record of something *working memory is a category of short-term memory (categories, not synonymous)

Hippocampus and all mammals

-the hippocampus is a masterpiece -evolved somewhere after species diverged and since the hippocampus looks relatively the same for all mammals means that it was conserved *hippocampus has not been updated in over 200 million years *why? -- goes back to the idea that the hippocampus is a masterpeice -take a slice of the hippocampus (slice like a banana) from a human, a tree shrew, and a rat and you will see that it basically looks the same on the inside

Atkinson-shiffrin cognitive model of short-term memory

-the multistore model of memory (also known as the modal model) was proposed by Atkinson and Shiffrin (1968) and is a structural model *this is a model of cognition and not of the brain they proposed that memory consisted of three stores: a sensory register, short-term memory (STM) and long-term memory (LTM) -information passes from store to store in a linear way, and has been described as an information processing model (like a computer) with an input, process and output -information is detected by the sense organs and enters the sensory memory -- if attended to this information enters the short term memory -information from the short-term memory is transferred to the long-term memory only if that information is rehearsed (I.e. repeated) -if maintenance rehearsal (repetition) does not occur, then information is forgotten, and lost from short term memory through the processes of displacement or decay -short-term sensory store = not quite short-term memory or long term-memory -sensory registers = tube and pipe for visual and auditory information -short-term store box = the 1980's version of this box can be expanded into the Baddeley's working model -long-term store = long-term memory and that box is a box portraying declarative memory (which they did not know or describe at the time)

Phrenology

-the theory that part of your brain is responsible for a certain topic, subject -the thought was the more you learn about that particular topic, that area of your brain would grow and you would feel a "bump" on your head because your brain is growing *"regional specificity"

Fear conditioning to a tone in rats

-training rats to be afraid of a tone -put rats in a box and x time after a tone rang, the rats would get a shock from the ground *response to the shock: rats jumped when they were shocked -the experiment tried to train the rats to be sccared when the heard the tone and they tested this by cueing the tone but not having a shock and observing rat's response *response: the rats would freeze during the tone (which is their fear repsonse) -contextual fear conditioning is the conditioning of a fearful situation to a previously benign context or location (rats starting to be afraid of a certain room) -animals condition both to the tone and to the environmental context in which tones and shock have been paired *this is selective to the exact environment in which conditioning occurs -amgydala lesions blocked fear conditioing to both tone and context, while hippocampus lesions did not block conditioned fear response to tone but did sig. with conditioned fear response to context

Fear conditioning to a context

-training the rats to be afraid of a sound with the rat jumping as their response to being shocked -except now testing if the rat is afraid of the room but placing the rat in the same room but without the cued tone or shock and seeing if rat freezes and is scared

***A working memory task for monkeys

-turns out monkeys have a prefrontal cortex - good agreement that there are regions in a monkey brain that could correlate to the ventral PFC -the last square is where the monkey should look where the black square appeared on the second screen once the fixation point disappears -what is going on in the prefrontal cortex during this delay period? *data from one neuron and it is data from one monkey doing the working memory task *all of the graphs show different trials of the neuron firing -this task seems relevant and significant enough that this task does in fact correlate to working memory *graph shows you increased neural activity during the cue and delay periods (Increased firing during cue and maintained it during the delay)

Working memory vs. long-term memory

-what is going on in the brain to support working memory is fundamentally different than all other mechanism that support that functioning of long-term memory so this class will focus on long-term memory

Profoundly severe amnesic patients H.M & E.P.

-white signal in the scan above represents cerebral spinal fluid and lack of brain tissue = similar amount of white signal in both scans (photo in camera roll) -E.P. brain damage was due to herpes viral infection and H.M. brain damage was due to surgeons removing those sections of his brain *EP has severe damage to the hippocampus and the area surrounding the hippocampus (completely lost hippocampus)

Working memory capacity relates to reading speed

-working memory seems to correlate not so much with the amount of digits there are but how much you can squeeze into a 4 second loop (so speed of reading matters) -capturing the ability of working memory seems to fit better into the idea of the phonological loop and not so much into a specific amount of numbers *aka this graph is showing that your capacity of working memory probably relates more to reading speed than it does to a specific number of digits you can remember

Amygdala and the physiological response to an emotional incident

-you LEARNED to have a fear association with bears and certain animals -what goes on in your brain: bear predicts bad things, which leads to activation of stress hormones from the adrenal glands (adrenaline, etc.) & these hormones come back to the amygdala as norepinephrine (gets released in the amygdala) *adrenaline gets released through vagus nerve & goes to nucleus of solitrary tract (in brain) -a little bit of cortisol release is good for your memory but extremes of this are bad ("too scared" can actually decrease memory of this event)

"The forgetting curve"

-you only remember about 20% of everything you learn (applies to declarative memory) -80% of everything you learn, you will forget at some point -forgetting is the fate of most of your memories -but maybe this isn't all bad because memory is selective and could you really store every piece of info you learn?

Reminiscing

-your brains did not evolve to reminisce, they evolved to remember things for survival *remembering facts are better and more conducive for survival

Episodic and semantic memory hypothesis 1

1. . Episodic memory depends on the MTL. Semantic memory does not depend on the MTL. -Semantic memory is just somewhere outside the medial temporal lobe, it can be acquired directly through the cortex outside the MTL. Episodic memory does directly depend on the MTL though. -HM t is given a implicit name-stem completion task to explore this hypothesis because he has almost no MTL -Asked last name of famous people and got all of them right - and these people only became famous after he had amnesia (and his hippocampus, etc. removed) -So they concluded that he could make new semantic memories without the MTL -Would you be impressed if you could only remember one name per decade? (similar to what HM was doing with this experiment) -Without MTL, you acquire almost no new facts - this is virtually nothing for over five decades compared to a normal individual -Manns does not really agree with this hypothesis Example for this hypothesis: -This is patient KC, black areas are showing where he had damage to his brain - it is all over the place and no real pattern (do not know where this damage is exactly) *some damage to MTL but widespread damage -Tried to teach him new "semantic facts" in a controlled lab setting -Teaching him three words sentences and then asking him the fill in the blanks of that sentence - goofy sentences intentionally so could test if he actually learned the fact or not: "dentist cured hiccups" *thought that he had semantic memory, so they concluded that semantic and episodic memory must be dissociated at some point in the brain -patient KC could learn new facts (I.e. had intact semantic memory), but he could not remember the testing sessions themselves (so had impaired episodic memory) Evidence against hypothesis 1: -patient EP has extreme diffuclty learning 3-word "facts" -EP could not use flexibly the "facts" that he did learn *did very poorly in synonym test of the 3-word experiment **suggesting that semantic memory probably does rely on MTL to some extent Additional evidence against 1: -pt. EP has not learned jack since becoming amnesic

***Amygdala round 2 main points

1. Amygdala is involved generally in emotional processing 2. We are considering its role in emotional memory 3. Amygdala plays two roles in emotional memory a. enhancing memory for emotion-inducing material b. attaching positive and negative dispositions to items 4. The cascade of physiological responses includes release of norepineprine in amygdala, which triggers amygdala's modulation of hippocampus (and, although we didn't talk about it, other memory areas) 5. Too much emotion or arousal can be bad 6. Although emotion DOES enhance memory under some circumstances, "flashbulb" memories are probably not entirely accurate

***Main points about the amygdala part 1***

1. Amygdala required for fear conditioning to either a single stimulus (e.g., tone) or a context 2. Hippocampus required for fear conditioning to a context but not to a tone 3. Plasticity in the amygdala itself may directly support acquisition of positive or negative dispositions to individual stimuli *reduction in fear of heights with drug that promotes synaptic plasticity (D-cycloserine) 4. Fear conditioning may be related to phobias 5. Extinction of fear responses (including phobias) represents new learning

Main points about cognitive models of short-term memory

1. Cognitive models were created without specific intention to model brain (biology/mechanisms) 2. Some cognitive models still useful as rough sketches of how brain works though 3. Other models have less direct reference to neuroscience

Three ideas about episodic and semantic memory

1. Episodic memory depends on the MTL. Semantic memory does not depend on the MTL. 2. Episodic memory depends specifically on the hippocampus. Semantic memory does not depend on the hippocampus. 3. Episodic memory and semantic memory both depend on the hippocampus and on the rest of the MTL. (This is the view held by your instructor.)

Main points about anatomy of the frontal cortex

1. Frontal cortex includes several areas, including some that are not directly relevant to working memory 2. Areas that are important are in prefrontal cortex: dorsolateral prefrontal cortex and ventrolateral prefrontal cortex (by prefrontal cortex we mean areas in the frontal lobe not involved with motor cortex)

Main points from hippocampus 2 slides

1. One distinction between MTL-dependent (declarative) memory and MTL-independent (nondeclarative) memory is the distinction between conscious and nonconscious memory. 2. This distinction is not easily applied to experimental animals. 3. Yet hippocampal memory system is similar across mammals 4. Other characteristics help define MTL-dependent memory (i.e., it tends to be rapidly acquired, flexibly expressed, relational in nature, and often includes remembering an item in its context). 5. The connectivity of the MTL with the neocortex (including the prefrontal cortex) helps us to understand why this memory system might operate the way it does.

Short-term sensory store main point

1. Short-term sensory store somewhat like after-image (or mental echo for sounds) 2. Sperling's partial report task helps measure short-term sensory store 3. Information "rescued" by being brought into working memory 4. We will not talk about neuroscience of short-term sensory store

***Main Points on the History of Memory

1. Some early philosophers had good ideas about memory but did not ground their ideas in empirical data (data/info acquired by experimentation or observation) 2. Darwin's ideas about evolution led to more empirical research in many areas, including psychology and neuroscience 3. One of the earliest questions about memory and the brain was whether memory was localized or distributed -- the answer is BOTH 4. Hebb's idea of "Cell Assembly" is how we will conceptualize memory 5. Behavioral psychology probably delayed the modern era of memory research 6. Patient HM marks the start of the modern era of memory research 7. Lomo and Bliss were first to study a neural mechanism of memory -- how a connection between two neurons could be strengthened

Neural correlates of working memory main points

1. Working memory is supported by persistent neural activity 2. That activity can be seen in prefrontal cortex (it is likely in other areas as well) -- dorsolateral and ventrolateral PFC 3. Different parts of prefrontal cortex may play different roles in working memory 4. Perhaps "working memory" is too broad of a psychological concept to accurately capture the various mechanisms supported by various parts of frontal cortex

Main points of working memory examples

1. Working memory should perhaps be thought of as "working with memory" 2. It includes: a. subvocal rehearsal b. active online maintenance c. chunking of info prior to storing in LTM d. processes after retrieval from LTM such as source memory and meta-memory (all parts, a-d, in #2 can be classified as "working with memory" instead of solely working memory)

MTL-dependent memory (declarative memory) tends to be:

1. rapidly acquired 2. flexibly expressed 3. relational 4. item-in-context

Episodic and semantic memory hypothesis 2

2. Episodic memory depends specifically on the hippocampus. Semantic memory does not depend on the hippocampus. (essentially same as 1st hypothesis but limiting region to the hipppocampus and not entire MTL) Example experiment: -did the test where you copy down an image and than are asked to re-draw that image x amount of time later from memory -did experiment with Beth, Jon, and Kate, three people with damage to hippocampus sustained before age 4 (but tested as young adults) *aka developmental amnesics --Beth, Jon, and Kate all graduated high school within a 1 or so years of their peers and had approximately normal IQ scores --were not able to copy down entire image but were able to put something on the paper (one of two shapes that vaguely represented shapes from the image) *so they were able to learn plenty about the world so doesn't this mean that they have developed new semantic memories *tested patients so young because 4 year olds do not know very much so they should not learn very much new stuff as they grow older if their hippocampus is essential to sustaining new important semantic memory --Rebuttal? *Since their brains were developing, maybe there was compensation and reorganization in the brain to take over this function

Overview of hypothesis 2 about episodic and sematic memory

2. Episodic memory depends specifically on the hippocampus. Semantic memory does not depend on the hippocampus. -some evidence for this idea -counterarguments exist for those data

***Hypothesis 3 about episodic and semantic memory

3. Episodic memory and semantic memory both depend on the hippocampus and on the rest of the MTL. (This is the view held by your instructor.) Experiment: -7 individuals with brain damage to the hippocampus did rey-o test *what one can see is that the delayed reproductions by the seven patients in the present experiments are more similar to the drawing by the patient with damage to the CA1 field than to the patient with widespread damage in the medial temporal lobe. All the patients remembered having drawn a figure previously and remembered at least that the figure was an abstract line drawing. Thus, the patients have a modestly severe memory impairment consistent with damage limited to the hippocampal region. Experiment 2 -asking patients with confirmed damage to hippocampus where or not someone is famous and if they recognize that person's name, is the person alive or dead -asking living vs. non-living question is because people die on a specific day and asking whether or not they died is directly testing memory (semantic memory) *only ask is person if alive or dead for the famous person that they patients DO KNOW AND RECOGNIZE -has to do with episodic and semantic memory because when someone dies, it can be an episodic memory but the fact if they are a famous person or not is semantic -controlled for episodic contamination by taking out data for people who said they had episodic memory of how they know when a famous person dies -patients with confirmed damage to hippocampus have not acquired a normal amount of knowledge about the world since sustaining damage *get better scores for people that they could have learned about before sustaining damage (similar to controls) and a lot worse of scores about people they could have learned only after damage *same data holds true even if benefit of episodic is removed

Overview of hypothesis 3 about episodic and sematic memory

3. Episodic memory and semantic memory both depend on the hippocampus and on the rest of the MTL. -patients with confirmed damage restricted to hippocampus are impaired in acquiring both episodic memory AND semantic memory

Amygdala's role in emotion

Amygdala plays many roles in emotion: -emotion regulation -emotion and social behavior -emotion and perception -***emotion and memory (WHICH IS KEY IN OUR CLASS)*** *non-declarative acquisition of positive and negative dispositions to stimuli *modulation of declarative memory for emotion-inducing material

Anatomy of hippocampus and MTL

And I want to point out that it's not just the hippocampus that is conserved, but many parts of the hippocampal memory system, including the adjacent cortical areas in the parahippocampal region that provide much of the cortical input to the hippocampus.

Why does activity in the hippocampus correlate with memory for emotional material?

Because although the amygdala is seen as the primary structure responsible for emotional memories, the hippocampus can modulate the amygdala. Increased activity in the amygdala correlates to increased emotional memory and the amygdala modulates the hippocampus, so this increased synaptic activty leads to stronger emotional memories. Emotional and declarative memories rely on the MTL, in which both the hippocampus and amygdala are located.

Circularity of memory definitions

Declarative memory is MTL-dependent memory and MTL-dependent memory is declarative memory....... and so on

***Anatomy of frontal cortex

Different areas of the frontal cortex important for working memory: -dorsolateral prefrontal cortex (dorsal means up/higher in brain) -ventrolateral prefrontal cortex (ventral means down/lower in the brain) -anterior rostral means towards the front of the brain, towards frontal lobe, while posterior caudal means towards the back of the brain towards the occipital lobe -pre-frontal cortex terminology: everything in front cortex that is not part of the motor cortex -dorsolateral and ventrolateral cortexes are important for "working with memory"

Overview of hypothesis 1 about episodic and sematic memory

Episodic memory depends on the MTL, while semantic memory does not depend on the MTL. -there is not much evidence for this (besides the general 3-word task) -unlikely that semantic memory can be acquired directly by neocortex (it is involved in higher functions such as sensory perception, generation of motor commands, spatial reasoning, conscious thought, and in humans, language)

Patient HM recap

H.M.: Henry Molaison -Originally studied by Brenda Milner (1957) -H.M. had medial temporal lobes removed bilaterally as last effort to curb epilepsy -Surgery led to memory impairment that is called anterograde amnesia -However, surgery left intact language, IQ, personality, perception, and working memory -H.M.'s case indicated that medial temporal lobes were important for memory -H.M.'s case was beginning of modern study of memory systems -HM could not encode new memories (lost some old memories too) *everything but declarative memory was still functioning *anterograde amnesia = cannot form new memories (no future memories) *retrograde amnesia = cannot access old memories -HM did not have "the encoding and retrieval from declarative memory via connections with MTL" neural network in place because damaged MTL

Source memory experiment example

Example of working memory: -showed two columns of words on a screen and then showed individual words on the screen and asked if that word was in the original list, and if so, which column was it in (left or right column) -recognition memory: an example of this is writing down yes or no about whether or not I saw the image with the word before -- not really working memory (more so declarative memory) -source memory: relates to asking us if we had seen that word in the original list and if so which column (L or R) was the word in -- asking us to work with (manipulate) our already retrieved memory -source memory is memory for the time and place in which an event occurred *source memory = taking individuals memories and connecting it to the source (remembering that bicycle was on the list and connecting it to which side of the page it was on) *sometimes called "working with memory" *source memory involves trying to link one memory with another -- connecting two different pieces of info

3-back task

Example of working memory: -this task involves listening to a list of numbers and raising your hand (or something like this) whenever you hear the same number in a list from 3 number places back (hence, the name of the experiment) *ex. 2 3 5 8 7 5 1 (number 5 in this list) -working memory goes beyond sub-vocal rehearsal -there are other ways to put info into working memory beyond sub-vocal rehearsal (counting, etc.)

Meta memory

Example of working memory: -using the same list of words with the two columns from the source memory example, he asked us to rank how sure we were that a word appeared on the list (ranking system is subjective and arbitrary) -what is the relationship between confidence and accuracy? *generally, on average (as tested in labs) confidence DOES correlate with accuracy *this example is not source memory -the meta memory part is the processing you do with memory - the specific example is confidence ratings but your confidence ratings are not memory (your Y and N are memories and your confidence about retrieved memory is you working with that memory so it is meta memory)

Patients with damage to frontal cortex & the 1-back and 2-back task

Experiment explained: -shows brains scans of patients with either ventromedial PFC damage, dorsolateral PFC damage, or ventrolateral and dorsolateral PFC damage *red blobs in the images is where the patients have brain damage -gave participants a working memory task (N back task) *did immediately adjacent computer screen have the same image as the original computer screen or did the one 2 screens after the original have the same image; tested same shape and same locations of shapes on screen Results: -2-back task is a harder working memory task to complete -what do you see in the 1-back task? *pretty much all the data looks around the same in both the shape and location tasks for all the groups -what do you see in the 2-back task? *it might not be so much about the shape or location, but something about 2-back versus 1-back and you get the most deficits when you have damage to both dorsal and ventral areas (patients with dorsal+ventral damage significantly do worse in both shape and location 2-back task -- a little worse in location though) Take-away point: -patients with damage to both dorsolateral and ventrolateral prefrontal cortex are impaired on a 2-back working memory task

What does the MTL consist of?

MTL includes hippocampus and adjacent cortical areas *such as the perirhinal cortex, entorhinal cortex, and parahippocampal cortex **all ventral to the hippocampus

Positive stimuli and memory

Study: -This was the first study to examine the relationship between a and em -In this study, 8 male Ss viewed 12 emotional video clips and 12 neutral video clips (each on different days, the order of which was counterbalanced) -PET scanning was done while watching both films -3 weeks after the 2nd video session, there was a surprise memory test in which the subjects were asked to recall as many of the clips as possible -rGMR was then correlated with memory performance for each individual Results: -Results showed that activity in the right amygdala was significantly related to the number of emotional film clips recalled, but it was unrelated to the number of neutral clips recalled -Additionally, if you look closely, there is activity in the PHC for both conditions -That is a pretty whopping correlation -Why only the right side? *Nobody knows. This is one of the more puzzling aspects of amygdala function: there are frequently found lateralized differences with no real explanation of why they would be that way. *A study published this year by Cahill, et al shows that when female subjects are used in the same paradigm, they show LEFT sided amygdala associated activity. *NOTE: it is easier to induce negative emotions in lab/artificial settings than positive emotions, but that does not mean that you remember negative events better than positive events* -it is about the arousal and release of hormones rather than negative or positive emotions *Watching a funny video up to 30 mins after learning words improved 1-week retention **both comedic and surgical videos had this effect

A study with an emotional vs. non-emotional twist

Study: -this study was trying to show that you do not need your amygdala to form a declarative memory necessarily but that you do need your amygdala to remember an emotional story better than a neutral story -told the participant two stories (with images): 1. Emotional-response story: A mother and son leave home one morning; see image of a boy with his feet severed off and reattached by surgeons; mother goes home from hospital 2. Neutral story: A mother and son leave home one morning; similarly, a picture of a boy with severed feet gets shown but pt. is told that the picture is just make-up artists stimulating an injury; mother goes home from hospital *again, this study was trying to show that you will remember the emotional version better than then neutral version that does not stimulate the release of hormones and arousal Results: -No memory enhancement for emotion-inducing part of story for patient with amygdala damage -for healthy pts. the emotional story is remembered better because your amygdala releases hormones and adrenaline -brain damaged patients still remember emotional story better so it does not have to do with all bran damage, just amygdala brain damage -bottom graph shows that SM does not have a problem with declarative memory but there is not boost on the graph when remembering emotionally charged story -damage to the amygdala is much more rare than patients with damage to the hippocampus

Semantic dementia

Task: -patients with focal lesions of neocortex can exhibit semantic dementia -color and object recognition task *give the patient images and ask which one is colored correctly **show a green pumpkin and a orange pumpkin and a green and orange lettuce *give the patient images and ask them which is the real animal **picture of monkey and than picture of monkey with elephant ears -if patients have damage for only one thing, than it is for the target unusual, suggesting that in mild semantic dementia know some semantic information about different categories of things (animals, vegetables, etc.) Results: -patients with mild semantic dementia, do worse on the pumpkin recognition task than the celery task (target unusual task), but patients with severe SD do worse on target-typical task (labeling pumpkin as green because it is a vegetable) -for the second task, object recognition, mild and severe SD were much more impaired for the target-unusual answers -target unusual object = a pumpkin because it is an orange veggie and most veggies are supposed to be "green"

Mirror-reversed text reading

Task: -similar to mirror-drawing task but had to say the word that popped up on the screen, but again the word was in a mirror so it made the words harder to read (they were backwards essentially) -the researchers wanted to test a cognitive skill and not a motor skill so this specifically test cognition because the words are different and are not drawing the same image over and over Results: -called this form of memory non-declarative because found out it was a different type of memory and that is could not have been declarative -found that amnesic patients improved with repetition on this mirror-word image task as well, even though never remember doing it before *so this helped prove that improvements were not just due to motor memory

Mirror drawing task

Task: -this is when you tell someone to trace over an image except they are looking at the image through a mirror (so it is flipped and this makes things harder) -follow-up experiment researcher did with patient HM - give patient a paper with a drawing on it and you try to trace over the image and cannot go outside the blue lines (in the slide in class, gives us the example of tracing a star) *CATCH: you are seeing the image in a mirror so you are having to do this in reverse of what your body is used to Results: -HM's performance improved with repetition, even though he never remembered that he has ever done this task before --the difference between us and HM is that patient HM does not remember ever doing this task but he could still accumulate learning and motor skills in the face of almost zero declarative memory for the task itself --got better everyday and picked up where he left off everyday but never remembered it -could this experiment just be muscle memory and not involved with actual memory? *this is what critics argued at the time

Short-term vs. working memory

The new term, working memory, is used now. It is an expanded definition of the short-term memory, including also the manipulation of the temporary stored information.

Short-term sensory memory

The term Sensory refers to the initial process of storing information that is perceived through our senses. It lasts for a subtle period and it is regularly replaced by new data, as our senses work continuously.

Activity in the human amygdala correlates with declarative memory performance for emotional material but not neutral material

This was the first study to examine the relationship between a and em -In this study, 8 male Ss viewed 12 emotional video clips and 12 neutral video clips (each on different days, the order of which was counterbalanced) -PET scanning was done while watching both films -3 weeks after the 2nd video session, there was a surprise memory test in which the subjects were asked to recall as many of the clips as possible -rGMR was then correlated with memory performance for each individual -Results showed that activity in the right amygdala was significantly related to the number of emotional film clips recalled, but it was unrelated to the number of neutral clips recalled -Additionally, if you look closely, there is activity in the PHC for both conditions -- that is a pretty whopping correlation. -Why only the right side? *Nobody knows. This is one of the more puzzling aspects of amygdala function: there are frequently found lateralized differences with no real explanation of why they would be that way. A study published this year by Cahill, et al shows that when female subjects are used in the same paradigm, they show LEFT sided amygdala associated activity. -when you see emotional information, you, on average, engage your amygdala -no correlation between memory and neutral material -but there is a correlation between how emotional that memory is (how much amygdala is engaged) and how much/ well you are going to remember that memory

***Amygdala anatomy

Where is it located in the brain? -the MAIN idea is that the amygdala is connected to many areas of the brain *can modulate striatum, PFC, MTL memory system, sensory neocortex, cerebellum, & HPA axis *involved in conditioned emotional learning

Digit span example

Working memory example: -on average, English speaking college students can do mean of 7 numbers with standard deviation of 2 (can remember on average 7 numbers, your working memory) -working memory is one of the first things to decline with age (youth of prefrontal cortex) -"chunking": one aspect of learning memory; putting words into concepts and remembering chunks of words that are now concepts instead of remembering each individual word *example, 1 2 3 4 5 6 7 8 9 (this string of numbers would describe chunking because you are not remembering each number but mainly the pattern and string of numbers)

Declarative memory depends on:

a. entorhinal cortex b. perirhinal cortex c. parahippocampal cortex d. hippocampus e. all of the above *E is the answer*

Anterograde Amnesia

is a loss of the ability to create new memories after the event that caused the amnesia, leading to a partial or complete inability to recall the recent past, while long-term memories from before the event remain intact