82
narrows the definition of the property of interest to what usually would be considered just an indicator of the relevant properties.
For similar reasons, the very idea of measurement uncertainty is ill fitting with
operationalism: If a measured property is by definition what is measured by a given
set of operations, what is there to be uncertain about? In particular, the idea that
there may be uncertainty in the definition of the measurand—i.e., definitional uncertainty, the “component of measurement uncertainty resulting from the finite amount
of detail in the definition of a measurand” (JCGM, 2012: 2.27; see Sect. 3.2.4)—is
irreconcilable with operationalism, due to its inability to separate what is measured
from how it is measured; one potential consequence of this is misconception that
what is true of the indication values is also automatically true of the measurand, “so
that, for instance, properties are presumed to induce a linear ordering of people
because [test scores] do” (Borsboom, 2006: p. 429).
For these reasons and others,
7
operationalism—at least in its original and radical
formulation—has been almost uniformly rejected as irreconcilable with general scientific practice and vocabulary (Bickhard, 2001; Green, 2001), even by Bridgman
himself (“I feel that I have created a Frankenstein, which has certainly got away
from me”, as he once lamented – Bridgman quoted in Frank, 1956: p. 75). The definition of in the International Vocabulary of Measurement (VIM)—
i.e., the “quantity intended to be measured” (JCGM, 2012: 2.3, emphasis
added)—acknowledges the possibility of a distinction between the property with
which the measuring instrument experimentally interacts and the property to which
the produced values are attributed, and thus appropriately emphasizes that a model
of the measurand is unavoidably present, though sometimes only implicitly, and that
it is only through such a model that the produced information is transferable. Thus
operationalism, in and of itself, is simply incapable of explaining the societal role
played by measurement, or of providing a justification for its acknowledged epistemic authority.
4.2.3 Representationalist perspectives on measurement
Many of the same sociohistorical forces that helped shape the operationalist perspective—in particular, logical positivism, and more generally the strongly empiricist and antirealist attitudes prevalent throughout the early-to-mid twentieth
century—also helped shape what came to be known as the representational theories
of measurement (RTM), which departed from and elaborated on Campbell’s famous
7 For a longer discussion of the history of operationalism, see, e.g., Chang (2019).
The operationalist stereotype: Measurement is a model-free, purely procedural empirical process, which only provides information about the property
with which the measuring instrument interacts.
4 Philosophical perspectives on measurement
narrows the definition of the property of interest to what usually would be considered just an indicator of the relevant properties.
For similar reasons, the very idea of measurement uncertainty is ill fitting with
operationalism: If a measured property is by definition what is measured by a given
set of operations, what is there to be uncertain about? In particular, the idea that
there may be uncertainty in the definition of the measurand—i.e., definitional uncertainty, the “component of measurement uncertainty resulting from the finite amount
of detail in the definition of a measurand” (JCGM, 2012: 2.27; see Sect. 3.2.4)—is
irreconcilable with operationalism, due to its inability to separate what is measured
from how it is measured; one potential consequence of this is misconception that
what is true of the indication values is also automatically true of the measurand, “so
that, for instance, properties are presumed to induce a linear ordering of people
because [test scores] do” (Borsboom, 2006: p. 429).
For these reasons and others,
7
operationalism—at least in its original and radical
formulation—has been almost uniformly rejected as irreconcilable with general scientific practice and vocabulary (Bickhard, 2001; Green, 2001), even by Bridgman
himself (“I feel that I have created a Frankenstein, which has certainly got away
from me”, as he once lamented – Bridgman quoted in Frank, 1956: p. 75). The definition of
i.e., the “quantity intended to be measured” (JCGM, 2012: 2.3, emphasis
added)—acknowledges the possibility of a distinction between the property with
which the measuring instrument experimentally interacts and the property to which
the produced values are attributed, and thus appropriately emphasizes that a model
of the measurand is unavoidably present, though sometimes only implicitly, and that
it is only through such a model that the produced information is transferable. Thus
operationalism, in and of itself, is simply incapable of explaining the societal role
played by measurement, or of providing a justification for its acknowledged epistemic authority.
4.2.3 Representationalist perspectives on measurement
Many of the same sociohistorical forces that helped shape the operationalist perspective—in particular, logical positivism, and more generally the strongly empiricist and antirealist attitudes prevalent throughout the early-to-mid twentieth
century—also helped shape what came to be known as the representational theories
of measurement (RTM), which departed from and elaborated on Campbell’s famous
7 For a longer discussion of the history of operationalism, see, e.g., Chang (2019).
The operationalist stereotype: Measurement is a model-free, purely procedural empirical process, which only provides information about the property
with which the measuring instrument interacts.
4 Philosophical perspectives on measurement