4.5 Ways of Knowing

4.5 Ways of Knowing

The potential for multiple methodologies for interpreting data, gaining information, and developing knowledge can also make knowledge sharing difficult. The various methodologies through which knowledge is produced and acquired are referred to as ways of knowing. Agencies and the individuals within them may have different ways of knowing because they have different underlying perspectives on the proper methods for ‘sense’- making. Oftentimes, different backgrounds and goals lead to different ways of knowing, such as the differences in vocabulary, technical expertise, analytical methods, and use of conceptual models that can arise from different types of education or training. These differences can create barriers to knowledge sharing if agencies and the individuals within them disagree on the appropriate methods to produce and interpret information.

One barrier that may be of particular importance for knowledge sharing occurs when agencies hold different perspectives on the most appropriate approach for interpreting complex information. Oftentimes multiple analysis and interpretation methods exist to make sense of information, especially if the data that built this information are describing a complex system. These differing methods for sense-making can result in dissimilar conclusions about the meaning of the information. Consequently, these divergences can cause a barrier to coordination if one of the agencies is unwilling to accept the other’s conclusions, especially if the conclusion was produced through methods with which they were unfamiliar or did not perceive to be accurate or legitimate. This barrier may be especially significant when the information will be used for decisions of high concern to the agencies involved or if the information will be subject to political scrutiny.

As an example, two common approaches to information interpretation that can cause knowledge-sharing barriers are the use of statistical models versus simulation models.⁶⁵ Statistical modeling involves the fitting of models to observational data and is designed to provide direct information on the relationships between different variables. This approach, however, does not provide information about the underlying processes that cause these relationships.⁶⁶ In contrast, simulation modeling is used to mathematically represent underlying causation processes. This approach may be useful for understanding why two data points have a certain relationship, but only if the theoretical basis for the model includes all critical processes.⁶⁵ Agencies may differ in which of the two methods they see as most appropriate for analyzing their data, and they might be unwilling to accept analysis results or other knowledge based on the use of a different model type.

Another potentially problematic barrier relates to knowledge sharing in the context of uncertainty. When there is uncertainty about the system, such as epistemic uncertainty or ambiguity regarding underlying concepts, the potential for multiple interpretations of data and information arises. This plurality of interpretations can make it difficult for agencies to develop shared or unified knowledge. In particular, having inadequate information can lead individuals with different methods of interpretation towards diverging conclusions, some of which may support certain goals of the agency better than others. In these circumstances, there is no objective knowledge, which can be problematic because subjective knowledge may be contested between agencies that are attempting to coordinate with a unified vision. Therefore, uncertainty may add to barriers regarding knowledge sharing because it can compound issues regarding multiple interpretation methods and different ways of knowing.^67,68