Throughout the history of thought countless words have been written concerning what knowledge is. There are innumerous proposals, from different philosophical precepts, that have attempted to answer this question. In this sense, if we review the literature on Cognitive Science and epistemology, we can figure out that there are several theoretical models that can meet the goal of offering an adequete definition of knowledge.

1. Classical epistemologic model

The proposal of classical epistemology advocates a definition of knowledge from the notions of belief, truth value and justification (or argument). In this sense, a person A knows that P if and only if it fulfills the following three conditions: (a) A believes that P, (b) P is true and (c) is justified in believing that P.

At the first glance, the classical epistemological proposal provides a solid base to approach the identification process and knowledge representation in the context of an organization. In this sense, in order to conclude that a person knows a concrete thing (has a concrete knowledge), we only have to verify that this person has a belief; the belief that coincides with this supposed knowledge we attribute to him, that what he thinks is true and that this person is justified in believing in it (that this attributed belief has to be reasoned, not arbitrary).

2. Dretske's informational model

Fred Dretske, the American philosopher, introduced Knowledge in informational terms in 1981. He provides, from his definition of informative content, a definition of knowledge in informative terms: K knows that s is F and only if K's belief that s is F is caused (or is causally sustained) by the information that s is F.

Within this definition must be understood the terms "belief caused by information" as that belief caused by the information contained in the fact that s is F.

In short, restoring the definition of informative content, so that K knows something, K should have information of that something with probability equal to 1, therefore, knowing that s is F requires not only certain information about s (an appropriate or sufficient quantity), but the information that s is F.

Two important theoretical benefits can be drawn from this Dreskian proposal on knowledge: 

The first of these benefits is found in the fact that this definition allows us to explain the possibility of transmission of knowledge: when a speaker K knows that s is F and, among other things, ii sincerely asserts that s is F, the listeners will come to know that s is F from what the speaker says (respecting the principle of the introduced copy in the previous section). This communicative fact is met, according to Drestkian definition of knowledge, if K knows that s is F from the information that s is F, and if the transmission of this information is done with an ambiguity equal to 0.

The second benefit is something beyond the possibility of transmission of knowledge. What this definition mainly pursues is to reach the goal of distancing from those clasical epistemological theories that had presented knowledge as a justified and true belief. Dretske replaces the necessity for the justification of belief with causality of information. He seeks, in making such a change, to overcome the problems usually presented by these classical theories (the paradoxes of Gettier and lottery), and also gets an adequate argument against the radical scepticism thesis.

Dretske defends himself from the thesis of radical skepticism (that supports the impossibility of knowledge) clearly distinguishing the conditions of an information source from what is called the conditions of an information channel. While a source generates information, the conditions of a channel, although it is crucial for the transmission of information, do not affect the information circulating within it. In this respect, the communication channel should be considered as a series of conditions which the sign depends on, that either does not generate (relevant) information, or only generates redundant information. In short, the channel offers no relevant alternatives to the source, and what makes an information channel to be ambiguous is its charateristics, not the suspicions that may or may not circulate information within it.

3. Floridi's semantic model

According to Floridi´s semantic approach (2005a, 2005b), knowledge is constituted in terms of justifiable semantic information, i.e. information constitutes the elements for further inquiry. At the same time, information is the result of a data modelling process. But unlike Dretske’s naturalistic assumption, this data modelling does not necessarily represent the intrinsic nature of the studied system, or it must not be directly related to the system by means of a causal chain; instead, it will depend on the processing of data by knowledge. In turn, data are conceived as the resources and restrictions allowing the construction of information. Therefore, it can be stated that Floridi proposes an architectural relationship between knowledge, information and data, being knowledge on the summit and data on the base. At the same time and as a result of such interrelationship, he replaces Dretske’s requirement of truth of (which is also subscribed by the situation theory) by a requirement of truthfulness, i.e. instead of searching for a correspondence between the statement and what the information is about, the attention is rather paid in the correspondence between what is reported and the informer.

4. Systemic model of the Unified Theory of Information (UTI)

From a detailed approach to system theory considering different self-organization levels (from self-restructuring to self-re-creation), knowledge is constituted in the →UTI by means of interpreting data (or meaning assignment) and is the basis for decision-making, which shapes “practical wisdom” (Hofkirchner 1999). 

UTI refers to different levels of information rather than dependency relationships, i.e. information is gradually processed: first, at the syntactic or structural level there is data, then at the semantic or state level there is knowledge, and, finally, at the pragmatic or behavioural level there is practical wisdom. The information processing is performed by means of interrelationship and reciprocal action between adjacent strata and not in terms of a casual progression (as in Dretske’s naturalism). In other terms, between micro- and macro-leves there are upwards- and down-wards causations (regarded as information processes) cooperating in the self-organizing processes.

5. Conductual model 

For example, it is argued that, relating to the conduct and actions of an agent, knowledge is the potential capacity that an actor poseses to act effectively. The effectiveness means to compare the behavior and potential outcomes with the objectives and values of both the actor and those of his community or the communities that he belongs to. 

Within this conceptual framework one argues that there are various types of knowledge. The first one is the knowledge of internal information. In this type of knowledge is the potential capacity of answering questions with correct answers; usually, the questions on real objectives, about the state of one part of the world in some time. For this kind of knowledge, it is a precondition that the actor answers without resorting to any external sources of information. Typically, any answers can be registered in records, which can be used by other actors. 

The second type of knowledge is knowledge of external information. This is like the previous one with the exception that in this case the access to other sources of information is permitted. 

In the third place, thinking is also a way of effective action. In this case, starting from available information, a process of creating new information takes place, which may become the answer to new questions or the spontaneous production of information by a thinking agent. 

Finally, there is a non-informative knowledge; the capacity of effective action is not related to information. It is something that one usually sees in artists and athletes. They can have a highly effective conduct most of the time, but they are unable to explain or articulate their knowledge on recorded information. 

6. Knowledge and near concepts 

a) Knowledge vs. information

From most points of view regarding information and knowledge, there are close relationships between these two concepts, especially as far as the common use of both terms is concerned. Usually, information occupies a lower position than knowledge, and the former –so to speak- ‘nourish’ the latter. However, this connection is disregarded in cases of a radical syntactic approach, in which the relationship question is avoided just addressing to the technical dimension (as in the MTC), or in a radical pragmatic approach in which only what-is-being-done is posed, that is, information is considered as a mere instrument of the action and, therefore, the problem of whether the information refers to states of affairs is ignored (either dealing with a correct apprehension or knowing that p is the case).

According to the above, the relationship between information and knowledge must evidently appear in all those informational approaches considering the semantic dimension, usually adopting a more analytic notion with respect to information, and a more synthetic one with respect to knowledge. Furthermore, a closer proximity to the object is used in information concerns, and to the subject in knowledge concerns.

For Dretske -as mentioned above-, "Knowledge is information-produced belief" (Dretske 1981, 91-92) and belief always relates to "a receiver's background knowledge" (pp. 80-81). From a naturalistic perspective, in which there is a casual dependence between the external conditions of a living being and and its internal states, information for Dretske creates experience (sensorial representations) and originates beliefs (cognitive experiences), which underlie the sedimentation of knowledge. 

b) Knowledge and mental states

We can agree, leaving aside the existing alternative definitions, that knowledge must be identified with a special kind of mental states (neuronal arrangements), presenting a set of particular characteristics, which an individual possesses. On the one hand, they are mental states achieved by the individual from a process of information assimilation or metabolism. This characteristic helps to distinguish those mental states of the subject corresponding to knowledge from those corresponding to mere beliefs, which do not reach the necessary epistemic level to be identified as knowledge.

In this sense, the semantic content of those mental states coincides with this assimilated information. And the mental states, conversely, act as a guide for actions and conduct of that individual; in other words, they control the decisions made by the subject. 

We can reflect this characterization in the following synthetic expresion: Knowledge = the mental states of an individual constructed from the assimilation of information, which steer the actions performed by the subject.

However, the characteristics of knowledge do not end here. We can elaborate a little more about this special kind of mental states. Knowledge, unlike data and information, is closely related to the actions and decisions of the subject, we can even evaluate this knowledge using as indicators such actions and decisions. Moreover, knowledge is the critical factor that permits the holder to assimilate new information -therefore, the creation of new knowledge-; and often it is continuously restructured by the entries of new assimilated information.

c) Knowledge vs. experience, truth, belief and values 

Nonetheless, it is not sufficient to provide a definition of knowledge and explain it with a couple of examples to have a better understaning of it. It is also necessary to deal with a number of related and interrelated concepts.

In this vein, we should not forget a concept very close  to knowledge, and which partly allows its acquisition: experience. Experience can be defined as the set of living-experiences that each individual has been through. And as such, it makes possible the creation of new knowledge through enabling the understanding of new situations from others that have been experienced, and to find new answers allowing us to adapt to new scenarios.

We should neither forget the concept of truth. As it has been defended since Classical Greece, knowledge (or at least a special type of knowledge, as we wil see) implies truth: if A (an individual) knows P, then it is true that P. If anyone knows that the water molecule consists of two hydrogen and one oxygen atoms, then it is true that this molecule presents the arrangement of atoms. And it is knowledge and its arising actions that have to be in tune with what really happens. Reality deals with refining and improving knowledge, rejecting and cleaning our heads from this supposed knowledge (pseudo-knowledge) that does not work and is not attuned.

Another closely related concept is belief, understood as the mental state that an individual possesses. Because knowledge (or at least one type of knowledge), besides truth, implies judgement or belief: in order that someone knows P, this someone has to believe that P. That is, knowledge must maintain a commitment to the truth of P. If someone knows that the water molecule consists of two atoms of hydrogen and one of oxygen, then that someone must believe that this molecule presents the arrangement of atoms.