This paper incorporates sociological concepts of knowledge, learning, and situated cognition. Text in red is commentary from the professor, David Moore.
Collaborative Work Software Concept
In 2008 I worked for six months as an executive assistant at a successful mid-size public relations firm on Park Avenue in midtown Manhattan. The experience convinced me of the horrors of the marketing world, but also gave me insight into how 40 disparate individuals can work together to create a cohesive company unit that can efficiently synthesize and distribute relevant knowledge. The firm generally had several clients at once, with several teams of individuals working on particular client projects (Teams were not unique and many had overlapping individuals. This tended to be a function of seniority—e.g. the president was involved to some extent in every project).
In this paper I argue for the use of a more systematic approach to working company knowledge, in which knowledge at each iterative level (project to employee to company) is seen as existing within a distributive framework. Distribution encompasses relationships such as: the division and reconciliation of knowledge and tasks among teams, access to knowledge according to hierarchy, and the interaction of knowledge with company policy and established social relations. This paper also assumes an operative definition of intelligence as defined by Gardner’s theory of multiple intelligences [Actually, I’m not sure why the concept of ‘intelligence’ would be necessary in an analysis of the distribution of knowledge in an organization – let’s see what you do with it.].
Knowledge Types: Overview
Many different activities occur during a typical workday at Park Avenue PR[1], each falls within two separate points of focus: generating new business or working for current clients. The former requires a sense of salesmanship, creative intuition into the unspecified desires and needs of a client, and a strategic approach to budgeting and project management. These items can be described using Gardner’s concept of multiple intelligences [Oh, I see --okay]. For example, salesmanship is a salient or common sense understanding of one’s interpersonal skills, in which an individual can advantageously perceive and react appropriately to others’ subtle communications of inner emotions and states. The creation of a mock project for a potential client draws heavily on this form of intelligence, while also utilizing visual-spatial or verbal-linguistic intelligences to different degrees.
The cumulative application of these types of intelligences by any one individual can contribute to the creation of a highly specialized set of knowledge, unique to that person. For example, Employee A may have a moderate level of interpersonal intelligence, and a high level of verbal-linguistic intelligence. Thus he makes a good candidate for drafting reports that communicate project outlines and goals to the client. Over time he will most likely work on projects that utilize his intelligence, thus garnering experience or knowledge specific to it. On the other hand, Employee B may have a high level of interpersonal intelligence, but only a low level of verbal-linguistic intelligence, making him the better candidate for a personal meeting with a client, in which the client relays goals and concerns. Each employee will most likely follow a different path of acquiring knowledge, one that is specific to the situations which utilize his knowledge [This vision of a distribution of labor driven by the intelligences of the various participants is intriguing – but overlooks the fact that organizations can’t (don’t) always have the luxury of assigning tasks that specifically. Often teams have to work on particular tasks, and the members are good at different things, to different degrees. Let’s see if you can correlate the individuals’ intelligences with the organizational structuring (distribution) of knowledge.] (this is an intuitive assumption, which presupposes that the average person will generally work in a sector or job that he or she is somehow suited for, as this is easier than working at a job in which he or she has to work relatively harder to produce average results).
Knowledge Relationships
Many of the employees at Park Avenue PR work on projects for the same client, or do similar work for different clients. However, each employee carries a unique knowledge set created from both the application of various intelligences and the experience garnered from working on client projects in unique combinations (i.e. not every employee works on the same pattern of projects or at the same time or for the same clients). Thus, there exists framework of knowledge within the company in the form of employees that can be viewed as analogous to a neural network. By exteriorizing and abstracting individual experience, one can define each network node as a client project or areas of knowledge-granting experience, while employees can be seen as the circuitry connecting these nodes in unique pathways [Interesting metaphor. But it may be backward: it implies that the employees are the transmission channel – the pipe – for the various forms of knowledge, while you have been suggesting that they are more like the repositories of knowledge, and pass it along (or perform it) at different points (or occasions) in the process.]. For example, Employee C has worked on five previous projects, three with healthcare companies and the other two with children’s clothing companies. She is connected to these five areas of experience based on the assumption that project participation equates to experience or knowledge of that experience. Now, if Park Avenue PR were to take on a pediatric center as a client, it would be advantageous to activate the knowledge pathway or circuitry of Employee C, as she has experience working with both children and healthcare. She also possesses the ability of the individual human brain to process and integrate these experiences, drawing informed conclusions that can be applied to novel situations. This is an important distinction as individuals differ from other distributed knowledge networks (such as companies) that have no central processing center from which to draw analysis [This analysis is intriguing (and courageous!), but I don’t think it quite works. The human brain doesn’t have a ‘central processing unit’ either. Vision, for example, doesn’t happen in any one section of the brain; the process is distributed across areas – and we don’t know where it’s integrated, though we know it is. One could see an organization as a similar distributed knowledge-processing function. And that may be what you’re asserting in the rest of this paragraph – with the (reasonable) claim that the organization needs ‘social acceptance’ and the brain doesn’t. . .or does it?!]. The argument here is that both employee and company can be seen as instances of distributed knowledge networks, but the knowledge contained by the company cannot be processed and instead becomes institutionalized according to social acceptance.
The company server is another excellent example of distributed knowledge as it provides access to documents that exist in a collaborative space. Anyone with access can alter the documents within a specific client folder (barring very specific restrictions to documents such as contracts), as well as read the information available there. An employee’s level of access to client folders serves the practical purpose of fostering collaboration among employee team members and maintaining client confidentiality, but level of access also implicitly builds a social hierarchy within the company based on knowledge [Does the social hierarchy stem from the level of access – or is it the other way around? Some people are more powerful, and thus have more access to knowledge. Or is it not a cause-effect relationship between social position and knowledge access?]. The CEO or president can access all of the company/client folders and thus has the ability to utilize all of the available knowledge, but she can also police the knowledge for collaboration and logic issues, or simply shoddy work. Low-level employees, having limited access to the company’s aggregate knowledge set, must instead collaborate with each other or more senior employees in order to respectively create or access what they might need. Interestingly enough, one’s access to the company server can act as a predictor of job stability. Intuitively, the more institutionalized knowledge an employee is granted, the more embedded he is in the institution. Likewise, the more of his knowledge that becomes institutionalized, the more knowledge nodes he contributes, and thus he extends the company’s available circuitry or network. Intuitively, this can be called employee value or worth.
The company server also acts as a collaborative space in which collective knowledge can be synthesized from the individual knowledge sets of each contributing team member. In this way, their knowledge sets can be institutionalized according to the social acceptance of other team members and the rest of the company. It is particularly important to emphasize the creation of new knowledge in this space as this is how new experience or knowledge nodes are created within the network (again, reinforcing the idea that employees are unique connection pathways between and among nodes [Again, the employees themselves are not the connection pathways; their actions in relation to others (and to technology) are the pathways: they have to do something to share their knowledge.]). Quite literally the server is an interactive map of the company’s knowledge network.
In conclusion, these are ideas I’m currently developing for the possible creation of an algorithmic approach to team building based on knowledge sets. Using a neural network analogy replaces a potentially biased heuristic of judging employee competency with a systematic approach to understanding the knowledge set of not just one employee, but an entire company. The implications for this analogy are numerous, and range from a quantifiable understanding of a company’s strengths (by analyzing the density of circuits within the network [i.e. do most circuits connect to similar nodes such as healthcare; if so a CEO can claim expertise in healthcare?]) to increased efficiency in choosing employee teams, or even predicting which teams will create a better product.
Sarah Henry:
This paper is really intriguing, and represents a serious effort to make sense of the distribution and use of knowledge in an organizational system. If I read you correctly, you are aiming at an ‘algorithm’ that shows an organization (a CEO, specifically) how to construct teams for varied projects based on the knowledge sets the individual participants have acquired through previous work experience. (Is that right?) I’m not confident that such a team-building practice can be algorithmic – there are always factors affecting the team’s efficacy that go beyond the specific knowledge sets of members – but the idea is interesting. I’m not sure yet of your neural network metaphor, either: although I actually like the idea that a group (I call it an activity system) can ‘have’ and use knowledge, and that it does function sort of like a brain, I don’t think we actually know enough about neural networks (at least I don’t – you may) to decide whether the analogy works. But it’s a creative and productive line of inquiry, for which I give you credit. You might have used some additional sources (there’s a good deal on distributed or shared cognition, for example), but this is a good start.
Grade: A- David Moore.
[1] Name change
