Instrumentation & Measurement Magazine 25-7 - 3

Measurement, Dataism and
Post-Truth Ideology: The Good,
The Bad and The Ugly
Luca Mari and Dario Petri
O
ur society is producing data in amounts orders
of magnitude larger than even in the recent past,
from many heterogeneous sources, often with no
or poor information on their quality [1]. In this unprecedented
situation, measurement, traditionally a privileged tool of
trustworthy information production [2], [3], is at a crossroads:
will it maintain and indeed enhance its societal role, or will it
become progressively irrelevant, losing its specificity? In this
complex and uncertain scenario, under discussion is the social
role of metrology itself-the " science of measurement and
its applications, " according to the International Vocabulary of
Metrology [4]. Some insights on this matter seem particularly
important today, when the more and more available big data
are inducing qualitative, and not only quantitative, changes
in the role attributed to data in decision-making processes
that pervade our daily life [5]. We propose here to understand
the current situation as induced by the tension between two
opposing ideologies: dataism and post-truth. After a short introduction
to such ideologies, we mention a few reasons why the
metrological culture, standing in between these two extremes,
could have a strategic role in a forthcoming knowledge-based
society.
Dataism
Dataism, at least in its most radical position, sees the universe
as a gigantic computation system, whose state transitions
are in fact " embedded " computations, so that empirical phenomena
are actually streams of data and living organisms are
biochemical data processing systems [6]. Hence data are the
objective representation of reality, or even the reality as such,
and therefore all that is required to understand the empirical
world and to make appropriate decisions on it. Producing and
sharing data is then the pragmatic principle of dataism: " If you
experience something, record it. If you record something, upload
it. If you upload something, share it " [7].
The so-called Data Science thus becomes the science of the
whole, a discipline that unifies all bodies of knowledge and allows
us to study such different phenomena as the harmonies
of a symphony, the structure of a financial bubble, and the evolution
of an epidemic disease by means of the same principles
and the same methods [6]. In essence, this is the unconditional
acceptance of the hypothesis that " data speak for themselves, "
aiming at purely bottom-up (because data-driven) descriptions,
predictions, and decisions that lead to " the end of
theory " [8].
It is a fact, that in some cases, descriptions, predictions,
and decisions can be effectively made from data only, and it is
this fact that could explain the widespread interest toward the
methods and the tools of Data Science. A good example is the
automatic translation between natural languages: its already
high and continuously improving quality, that a few decades
ago expert systems based on explicit linguistic models failed
to achieve, has been obtained by means of artificial neural networks
trained on large collections of multilingual texts. This
could be interpreted as a confirmation of the validity of dataism
itself: outcomes, that in the past were supposed to require
" opening the box " to study and understand how data are processed,
are now obtained by " leaving the box closed " and
operating according to an input-output, behavioral approach.
For sure, this loses the possibility to provide causal explanations
of the predictions that the system produces, whose
reliability is only a matter of goodness of data fitting. But for
dataists this is a further reason of advancement: they would
claim that causality is an unclear, unscientific concept [9], and
removing it from technical endeavors should be considered a
significant step forward.
Signs of the success of dataism can be found not only in
the diffusion of Machine Learning applications, but also in
the increasing reliance on formal and quantitative criteria in
decision making, as in the case of the so-called bibliometric
A first version of this paper, " La cultura metrologica tra dataismo e post-verità " (Italian), was published in the
Dec. 2021 issue of Tutto_Misure, a journal published by the Italian Association Gruppo Misure Elettriche ed Elettroniche.
October 2022
IEEE Instrumentation & Measurement Magazine
1094-6969/22/$25.00©2022IEEE
3
Instrumentation & Measurement Magazine 25-7

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 25-7
