Transactive Knowledge

Keywords: Sensemaking, distributed cognition, transactive knowledge, knowledge management.

Daniel Wegner -- Transactive Knowledge

Daniel Wegner, at the U. of Virginia, has developed a theory of transactive knowledge. Transactive knowledge means knowledge that is somehow available or possible because of transactions between people. Here is how it goes.

Internal & External Memory

How does one retrieve a memory? color of a tomato. we know billions of facts. should take time to look that up. but it doesn't. It is stored with the word tomato. It's connected. Thinking of tomato does not bring up waterfowl. Not connected. It's a mental -- neural -- network. This is how our internal memories work.

How does one use external memory? We write down appointments, books, etc. what do you need to retrieve external memory? A label or cue. say you are looking up rudy's number. You don't just look at all the phone numbers in your book and then recognize it -- you might not recognize it all. But you can find the word Rudy, and then pick up the number next to it. So "Rudy" is a label or retrieval cue. 

You also need a rough idea of where the memory is stored -- as in, an address book. The same is true for computers. For a computer to find something, it needs a label and a rough idea of where to look. 

Computer Memory

Computers can also share memories. Sometimes the memories are tied together in the sense that all the computers use the same memory space. Human being can't do that. Computers can also share memories by having separate memory spaces, but sending messages to each other requesting various bits of data. To do this, they need to have knowledge of what's in the other computer. They need a kind of table of contents or rough version of what's in the other's memory, as well as a label for finding it and a rough idea of where it is.

Human memory sharing

Let's look at human sharing of memory. Without ESP, we do it largely through external storage media, like books and notes. Sometimes the external memory is a person. Marge, what's bill's number? Where are the candles? 

Sometimes, another person is used, but not because they know. 

"Jean, where is the salary report?" 
"I gave it to you tuesday." 
"Oh yeah. ok. I think it's on my file cabinet."

Here, it's the interaction of the two people that retrieves the memory.

Responsibility for Information

People are often assigned responsibility for information sharing, and people accept the responsibility.

Human groups readily associate certain individuals with certain domains of expertise, and vice versa. That is, they don't all know the same stuff. They keep the expert on x informed about developments in that area and then forget the stuff themselves. It relaxes people to know that they have gotten the info to the right person, and now they can forget it.

And the last person to deal with some domain sometimes is held accountable for it. The last person to use mower is now somehow responsible for information about it. "Bill, how's the rust on the mower?" "How much gas is left?". First person to mention that some band will be in town, is kind of expected to know more about it, and will be asked questions about it.

Formal groups, such as organizations, frequently assign people to go find out about something or generally be in charge of certain areas of information.

So part of what makes transactive memory possible is that people do take responsibility for knowledge.

Couples Research

Couples also do this. Both members of couple can be exposed to something, But, if it's in the other person's area of control, the one person lets it pass, while the expert "catches" the item before it is lost to the couple. Experiments show that it is hard for respondents to learn material in areas that they know the partner is expert in. In the experiments, when something is made to interfere with the knowledge of who is the expert (e.g., by assigning expertise roles in the experiment that differ from the natural roles), stuff falls through the cracks and the pair forgets things.

Experiments also show that if you make one partner responsible, they will learn better than the other, but only if they believe the material is not something the partner is expert in.

In the dissolution of relationships, one issue is that memory capacity is suddenly halved and both individuals become temporarily incompetent, leading to depression, low self-esteem, etc.

Group Retrieval

There is no guarantee that information held by an individual can be retrieved in its original form. Not because the individual changes it -- that's always true, but because in the group setting retrieval means talking about it, and when people talk about stuff, it changes.

People also talk about stuff when the information is being encoded. When that happens, labels are linked to the item by the conversants as a group. This makes it easier to retrieve the info because everyone knows the labels under which the information was stored. But it also can cause problems. For example, whatever label is applied first, perhaps arbitrarily or mistakenly, becomes the catchword for the item. It becomes a common denominator for discussion and becomes part of the item for everyone. Then it's hard to change.

On the hand, talk after information has been encoded can also cause changes. UFO sighting, person comes home tells all about it, goes to bed. The people who didn't see it talk about, resolve inconsistencies, end up with majority view that differs from the original story. Because it's a group, they may later convince original witness that they are remembering it wrong.

[Even talk about the opposite can forever associate something with something else. Nixon's campaign. Some people say so and so is a communist because ... I'm not one of them. I certainly disagree with her, etc. etc. Again, people say she's a communist. .... At the end, everybody's wondering if she's a communist.]

When multiple people are held to be in charge of some information, it often gets lost because each thinks the other's got it.

Often, the person held to be in charge of some information is the first person to handle it. Like whoever picks up the phone is responsible for getting the message to the right person. In organizations, the phone handler is often a secretary who doesn't know what each person knows or does. Is easy for them to send things on to the wrong person.

One key to transactive memory is the labels that cue retrieval. In intimate couples, the labels become idiosyncratic, understood only by each other.

Another key is the development of an accurate understanding of what the other's expertise is.

A third key is the differentiation and specialization of functions. Usually things evolve so the fewest people that can handle a function will carry the load, and responsibility is continuous over time rather than intermittent. This may account for some sex role differentiation. Females may have headstart for various reasons on baby information, and this kernel of expertise then draws additional information from others, and soon she is the only one with any knowledge of the baby.

Fourth, in sharing lower-order, detailed, disparate information, they often discover higher order themes, generalization and ideas that subsume these details. These integrations of disparate inputs that seem to embody the "magical transformation" that group mind theorists seek to understand.

Fifth, as mentioned before, is the issue of people accepting responsibility for information.


Health Behavior

People usually diagnose themselves -- at least enough to decide whether to go to drugstore or doctor and which kind of doctor. But actually it's the family and friends and coworkers that do it. They provide prototypes of illnesses (if you have a lump it means x; when you have mono, you are tired all the time). Circles of friends who have relatively impoverished cognitive prototypes lead to individuals making mistaken or late diagnoses. So your health is partly a function of the knowledge of our personal networks.

The doctor-patient dialog is also a transactive memory system. They have to work together to figure out what's going on because the doctor knows about illnesses and the patient knows about self and symptoms. Both are needed for correct diagnosis. Then there is the prescription, which can be complicated, which usually involves the whole family.


Think of aircraft carrier operations. Here is an informant's description (reported by Rochlin, LaPorte and Roberts (1987):

... imagine that it's a busy day and you shrink san francisco airport to only one short runway and one ramp and one gate. make planes take off and land at the same time, at half the present time interval, rock the runway from side to side, and require that everyone who leaves in the morning returns that same day. Make sure the equipment is so close to the edge of the envelope that it's fragile. Then turn off the radar to avoid detection, impose strict controls on radios, fuel the aircraft in place with their engines running, put an enemy in the air, and scatter live bombs and rockets around. Now wet the whole thing down with sea water and oil and man it with 20 year olds, half of whom have never seen an airplane close up. Oh and by the way, try not to kill anyone."

This is a million accidents waiting to happen, yet few do. Why? Systems that absolutely need reliability are organized a bit differently. They are organized so that things don't slip through through the cracks, so that individuals do very simply tasks, but which together result in coordinated action that looks like a giant mind was running it all without forgetting a single step.


Brain model in which knowledge resides in patterns of connections among vast numbers of fairly simple processing units. Each unit's activity is regulated by the activity of neighboring units. But it takes a lot of units. The human population on the earth perhaps achieves this.

What connectionism contributes to organizational theory is the insight that complex patterns can be encoded by pattersn of activation and inhibition if the units are richly connected.

Hutchin's (1990) description of coordination in naval navigation teams:

... the sequence of action to be taken need not be explicitly represented anywhere in the system. If participants know how to coordinate their activities with the technologies and people with which they interact, the global structure of the task performance will emerge from the local interactions of the members. The structure of the activities of the group is determined by a set of local computations rather than by the implementation of the sort of global plan that appears in the solo performer's procedure.

Robot walking systems -- imitate ants. Birds flocking. How to achieve coordinated action using local computations. It works by interconnection. Mutual adjustment.

Narrative skills are important for collective mind because stories organize know-how, tacit knowledge, nuance, sequence, multiple causation, means-end relations, and consequences into a memorable plot.