Knowledge Uniqueness

Knowledge Hierarchy

Many scholars conceptualize a hierarchy of data, information and knowledge. As illustrated in Figure 1, each level of the Knowledge Hierarchy builds upon the one below. For instance, data can reduce uncertainty or equivocality, and they are required to produce information, but information involves more than just data: the data must be in context in order to inform (e.g., so someone can ascribe meaning to a message). Similarly, information can help people make sense of their environments, and it is required to produce knowledge, but knowledge involves more than just information: it enables direct action (e.g., good decisions, appropriate behaviors, useful work; yes, judgment and norms determine what constitutes “good,” “appropriate” and “useful”). These three concepts are clearly complementary, yet they are clearly distinct also.

Figure 1.

Knowledge hierarchy (adapted from Nissen, 2006)

Consider this Gedanken experiment. Say someone sends you an e-mail message. This person prepares the message meticulously to ensure it is correct factually. It is sent via Internet and arrives in your computer mailbox within seconds of being sent. The message suffers from no transmission errors and appears on your screen exactly as it was sent. The message is written in Korean. Is this knowledge (e.g., reading the message enables you to take direct action)? Is it information (e.g., you can ascribe meaning to the message content)? Is it even data (e.g., it reduces uncertainty or equivocality)? Unless you can read and understand Korean, the answer is probably “no.” In such case, you would have received visual signals (e.g., symbols on the display screen) that do not even represent data to you. Many otherwise capable intelligence analysts confront this situation daily when searching for clues through communications (e.g., e-mail messages, recorded telephone conversations). Notice here that some knowledge (e.g., of Korean language) would be required to interpret signals into data.

Now say the message is written in a language you can understand (e.g., English). The message is: “333/33.” Is this knowledge, information, data, or simply signals? Unlike the example above, in which a lack of language knowledge precludes even the interpretation of signals from conversations into data, here no such language barrier exists. One can say the message provides data: a person can interpret the signal as a compound symbol with three “3”s preceding a “/” and two more “3”s, but can we say this data symbol represents information or knowledge also? Unless the symbol comes in the context of an answer to a factual question (e.g., how can the repeating decimal number 10.09 be represented by two whole numbers, with five digits in total, and both ending with the digit “3”?) or other, comparable context (e.g., mathematical symbols), the answer is probably “no.” The symbol constitutes data but not information or knowledge. In other words, the symbol can reduce uncertainty or equivocality, if the appropriate context is known, but it is unlikely that such symbol alone would enable someone to ascribe meaning to the message (i.e., constitute information). Notice here that some knowledge (e.g., how to place the message into appropriate context) would be required for data in the message to inform its recipient (i.e., for information to flow). Notice too that such contextual knowledge (e.g., mathematical symbols) differs from the kind of language knowledge (e.g., Korean) from above that supports signal interpretation. One should get a sense that different kinds of knowledge (e.g., language, contextual) are required for different kinds of actions (e.g., interpretation, informing) related to the knowledge hierarchy.