About Long-Term Memory Codes
This is a series of blog posts about the learnings and discoveries I made while studying Cognitive Psychology at the University of Berkeley, California. It will introduce you to the basic concepts of human cognition; attention, memory, categorization, semantic organization, language, problem solving and decision making.
A introductory side story
It was about seven years ago when I finished my last tour as an eco-exploration guide for an outdoor tour operator in Whistler, BC, Canada. I do have a vivid and very detailed (episodic) memory of some of the key moments and events of my life as a professional guide. What I seem to have completely lost, or currently are unable to retrieve, is the information about the eco-curriculum we delivered on each tour. It was about six pages long and I remember using different knowledge acquisition techniques (control processes), such as verbal rehearsal, coding and imaging to transfer it from my Short Term Memory (STM) into the Long Term Memory (LTM). That was in April 2008, just shortly after I applied and got the job with Ziptrek EcoTours. Like Atkinson & Shiffrin, although for different reasons, I focused most of my study time on verbal rehearsal because at this stage, my English was just not good enough and I needed to rapidly memorise many facts in a new, unfamiliar language. I even had native English speakers, translating some of my German notes into proper English sentences. You saw me speaking out loud newly acquired English words and sentences, over and over again. Short stories about the local environment, the flora and fauna, until finally, I managed to recite all six pages from my long term memory. For the next three-and-a-half years, I conducted about ~2000 individual tours and recited and taught roughly the same amount of the acquired eco-curriculum, with some variations and alteration over time. So, why is it now so challenging to retrieve all the information, after just 7 years and so many repetitions?
My assumption is, that it was because of my unfamiliarity with the English language, that led to the inability to encode the information effectively. What do I mean by that? Well, if we look at Squire’s (1994) taxonomy of the LTM, I believe that the facts of the eco-curriculum got stored in my Explicit Memory and the newly acquired English words got stored in the Implicit Memory, but because my English language skills were so rudimentary, the two separate memory components were not able to draw meaningful and effective links with each other (was the language acquisition to automated and unconscious?). I speculate, these ‘missing links’ are the cause of my bad retrieval performance for the eco-curriculum.
On Long-Term Memory
We learned in previous chapters (‘Cognition’, Stephen K. Reed, 2012), that our Long Term Memory differs quite a bit, from our Sensory and Short Term Memory and as Christopher Gade (University of Berkeley, 2017) in one of his videos suggested, some of these assumptions can or has been tested and confirmed already. A fascinating fact for me to learn was that our capacity in LTM is infinite and the information acquired can be very detailed. Although it is a bit of a challenge to get new stuff into our LTM, once it is in these areas of our brain, they really seem to stick. I used the word ‘area’ to highlight the fact, that LTM is not just one component or system of our brain, but exists of many (memory) components that work with each other and store different types or sets of information.
We can truly see the separation of multiple memory systems when we observe how people with amnesia perform on memory tests. Reed mentioned in his book the study of Warrington & Weiskrantz (1970) which unveiled, that the amnesic subjects did worse than healthy participants in Recognition and Recall Tests (Direct Memory Test), but did similar in Indirect Memory Tests (word-fragment and initial letters test). It seems, the damaged parts of their brain, were only responsible for certain memory types and info-snippets, but not for others. What truly helped me to understand the taxonomy of the LTM and it’s associated brain structures in more detail was Squire’s sketch (1994) in Reeds book (P. 120). There, he outlined all the different components of the LTM and where they are located in our brain, including what types of information, gets stored in which area. It is probably important to understand that our brain stores semantic and episodic information (explicit) differently, as well as implicitly acquired information or knowledge.
So far, I have only talked about some of the key characteristics of LTM. We still need to investigate, A) how do we get information into LTM (encoding), B) when do we know it is there, C) how can we retain it (storage) and D) how can we effectively retrieve information again (retrieval)? These are the three aspects of learning, shown in Nelson & Narens diagram from 1990.
The Serial Position Effect (Atkinson & Shiffrin, 1974) nicely demonstrates when information enters LTM, respectively if it’s still stored in STM. I can recall Chris Gade’s experiment that made us remember a list of twenty words, which were presented in a five-second delay. In between the presentation of the words, we were asked to repeat or subvocalize the words given to us and finally write them down on paper. A common pattern emerged, which is often referred to as the Serial Position Effect: We can recall the words from the beginning (primacy effect) and end better (recency effect), than the words presented in the middle. The recency effect shows us, that the words we recall are the words from our STM, and the primacy effect shows, that those words must be stored in LTM already. This can be tested, when we eliminate the recency effect with a retroactive interference task (min 30 seconds), just shortly after the presentation of the word chain (Postman & Phillips, 1965). Knowing about such experiments and the results of it, help learners to better determine (Metacognition) if knowledge has been properly acquired and — respectively — ended up in LTM. It provides inexperienced and experienced students like myself, the chance to increase awareness around variables (time, difficulty, judgement, retention interval, etc.) that heavily influence knowledge acquisition and retention.
We are now coming to the last aspect of learning, which is the retrieval of previously stored information in LTM. I admit, that I was taken by the stories and experiments of Elizabeth Loftus, which were first mentioned in Chris Gades second video of this lecture. I had to go online and look up more videos that dealt with Eyewitness Testimony and found a talk about ‘The Misinformation Effect’ and ‘Rich False Memories’ (link). I’ll get to that later. As a summary of this part, we can argue, that retrieving information from LTM is not an easy task, nor can it be done always with full awareness or extremely systematically like a computer program or Search Engine. We might have developed certain strategies to ‘search’ our memory effectively, leveraging cues or partial information like the first letter of a word or using (word/image) associations and drawing contextual links such as time, location, word associations and such. Sometimes we find ourselves in a situation where we know it is in our LTM, but we can’t retrieve it. This is the Tip of The Tongue phenomena.
There are other situations, where we start to alter the information that’s stored in our LTM. Reed and Gade mentioned a couple of powerful examples where alterations of storage took place, while people were asked to describe an event they had witnessed. Often leading questions or manipulations of certain contextual material like photos directed witnesses to inconsistent answers or alterations. Cognitive Psychologists in Canada were even able to demonstrate, that people admit to a crime they’ve never committed (link). Loftus talks about post event activity and how it can produce memory distortion. Right at this moment, I probably alter some of the information I read in Reeds book, or heard in Christopher Gades videos, mixing in new information from other sources or having mentally formed new models, about certain concepts that better match what already is in my memory. Sometimes, the information that eventually ended up in LTM was wrong in the first place because we’ve just decided that this is or was our truth — even though, it didn’t match the reality. Donald Trump demonstrates extremely well, that perception is selective.
We just need to be aware that recall and recognition are two different things we can also test for, as already mentioned in the first paragraph (Direct vs. Indirect Memory Tests, ‘Cognition’, Stephen K. Reed, 2012). For recall — knowing the context is crucial, which made me try a few of the four retrieval techniques mentioned in the book (reinstate context, repeat everything, recall events in diff. Orders, recall from diff. perspectives), but could not get far enough with recovering all the missing information of my eco-curriculum.

