Modification:Sciento-2019-0018

Accept that the relationship of tool reliance can obtain between epistemic agents and epistemic tools. Also accept the definition of tool reliance, which states that an epistemic agent is said to rely on an epistemic tool when there is a procedure through which the tool can provide an acceptable source of knowledge for answering some question under the employed method of that agent.

The modification was suggested to Scientonomy community by Paul Patton on 26 December 2019.1 The modification was accepted on 23 February 2024.

Preamble

The role of tools and instruments in scientific change is clearly important, but has not been defined within scientonomy. The question of whether epistemic tools can be subjects of authority delegation has been raised. This is not possible. The scientific data produced by some epistemic tools are a kind of theory. Theories must be assessed, in accordance with an employed method. This assessment requires semantic understanding of the data, and of the theories specifying the conditions under which the epistemic tool can be deemed a reliable source of knowledge. 1 Given the current state of artificial intelligence, epistemic tools lack the needed cognitive abilities to possess this understanding. 1 Therefore a new role for epistemic tools and their relationship to epistemic agents must be specified. The role proposed here, that of tool reliance, plausibly locates its basis in the concrete requirements of the employed method of some epistemic agent.

Modification

Accept that the relationship of tool reliance can obtain between epistemic agents and epistemic tools.

Also accept that the following symbol is to be used in scientonomic diagrams to depict the relation of tool reliance:

Theories To Accept

• Tool Reliance (Patton-2019): An epistemic agent is said to rely on an epistemic tool iff there is a procedure through which the tool can provide an acceptable source of knowledge for answering some question under the employed method of that agent.

• Epistemic Agent Can Rely on Epistemic Tools (Patton-2019): An epistemic agent can rely on any number of epistemic tools, while an epistemic tool can be relied on by one-to-many agent.
• Tool Reliance Exists: There is such a thing as tool reliance.

Questions Answered

This modification attempts to answer the following question(s):

• Tool Reliance: What is tool reliance? How should it be defined?
• Associations of Epistemic Agent: How is the class of epistemic agent associated with other classes (and itself)? What aggregation, composition, or other association relations can exist between epistemic agents, as well as between an epistemic agent and instances of other classes?
• Associations of Epistemic Tool: How is the class of epistemic tool associated with other classes (and itself)? What aggregation, composition, or other association relations can exist between epistemic tools, as well as between an epistemic tool and instances of other classes?
• Existence of Tool Reliance: Does tool reliance exist?

Verdict

The modification was accepted on 23 February 2024. The modification was discussed during the 2024 workshop. Firstly, the similarity between this definition and the recently accepted modification suggesting a definition of epistemic tool was highlighted by the community as an indicator that this modification might require further work to separate out its definitional content from dynamics content before it could be accepted. However, most of the discussion surrounding the modification concerned, as Jamie Shaw put it, whether the distinction between authority delegation and tool reliance was a distinction without a difference. Deivide Garcia suggested that there is a meaningful difference between the two: epistemic tools are inherently passive, whereas epistemic agents inherently play an active role both in tool reliance and in authority delegation. Paul Patton highlighted several examples of “troubleshooting” epistemic tools in the history of science, which is suggestive of a relationship between the epistemic agent and the epistemic tool that differs significantly from the relationship between two epistemic agents in an authority delegation relation. Yet, Hakob Barseghyan highlighted that the process of troubleshooting of tools could be conceptualized in different terms very similar to how scientonomers already describe authority delegation relations. Joshua Allen suggested that the modification would be useful to accept anticipating further work being done on the relationship between epistemic tools and epistemic actions. Ultimately, after most participants agreed that the modification was useful in its current state, the modification was accepted with over 2/3rds majority (11 votes to accept, 4 votes to keep the modification open).

Click on the Discussion tab for comments.