Making Time: A Study in the Epistemology of Measurement
- How is the second defined? What do standard clocks measure?
As a preface to answering this question, we must first understand what "the Epistemology of Measurement" means. According to Tal, it can loosely be defined as what ties the important the knowledge claims of abstanct ideas, such as time together with concrete forms of measurement, such as a clock (2016, 299). In other words, it deals with the assumptions, inferences and justifications behind measurement and knowledge.
Off of that, Tal explains that there are many claims and criteria for measurements to work properly. For example, the congruence claim is able to define a unit through equality: "magnitudes of certain quantities such as length and duration are deemed equal to one another" (2016, 311). In order to test this with reproducibility; that is, where two separate people would reach the same conclusion, scientists must be able to "standardize" the units.
Standardizing a unit, however, has proven to be difficult. This yields to the idea of conventionalism: where the choice of the congruence criterion is arbitrary, but all people would agree upon it, and it would work and be reproducable. (2016, 306) This leads to Tal's argument and viewpoint, where he draws on lessons from metrologists.
Conventionalism is static, but according to him with his model-based account of measurement, this should not be the case – the models do not need to be fixed and should be continually adjusted to ensure reproducibility. The criteria ultimately depends on how metrologists model their instruments theoretically and statistically (2016, 328).
With the above as context, I believe I can then answer the questions appropriately.
The conventionalist solution is universal coordinated time. It is the agreed upon time, the time that all of our devices are calibrated to. How does this work though? We need to ensure that the second is uniform and accurate across the globe, and currently, the theoretical solution is a definition of a second which is the duration of exactly 9,192,631,770 periods, or Hz, of the radiation corresponding to the transition of cesium-133 atom in the ground state (2016, 301). Further, we simply define the second at the sea-level (clocks tick at different rates over time depending on the gravitational potential energy). This is the conventionalist solution (and the current one), but the problem is that this is an abstract definition which is difficult to accurately calibrate to exactly. The assumption is that a cesium atom is not moving, but we know this is not the case because you can never actually reach absolute zero.
Because of this, different clocks measure different times since they depart from ideal conditions in different ways, with tradeoffs of stability and/or accuracy. Standard clocks, therefore, measure the "most accurate" approximation that has the least deviation from the standard unit as possible, such as the NIST-F1, although not perfect and cannot run as a continuous clock, and to assess how accurate a clock is, you would need to compare it to another, more stable clock, but this would eventually pull all of the clocks towards a slightly different frequency.
After answering the given questions, it yields the important question, is he right? He believes that a better approach is to create a hierarchy of models and parameters that allows us to mediate all of the levels, which creates a more accurate approximation. This means that it would go from the local levels to the coordinated universal time to the terrestrial time. Coordination is multi-step, not one, which means that we are creating ideal parameters from concrete instruments. This would allow us to use have reproducibility, it would not be static or arbitrary since they would have legislative freedom (2016, 322). I agree with him, I do not think conventionalism is right because it is fixed and would deviate from the right value in the long run, and it gives metrologists the freedom to adjust their models. But that ultimately is up to the individual to decide.
Comments
Post a Comment