Guest Essay

Metrology is hard enough - but are we sure we're speaking the same language when we discuss it?

The terminology of metrology: some reflections, with particular reference to the case of medical laboratories

April 2024
Luca Mari, Marco Pradella

(A previous version of this paper was published, in Italian, in the journal Tutto_Misure, 4, 2023.)

We know how to communicate in different ways, for example but means of gestures and images (as it is the case of traffic signs), but words are the best tool we have to this purpose. Through words, sometimes in the form of symbols and symbolic expressions, such as ‘H20’ and ‘2+2’, we customarily communicate, confident that we understand each other and thus that communication is successful. But is this trust justified? How can we know that we really understand each other when we communicate? How can we, who are writing these words, guarantee that you, who are reading these words, understand what we mean by writing them?

An extreme and simple case is communication in a language either of the speakers does not know: certainly, it is not successful. If instead we are speaking to refer to material objects in front of us, and we do so in a language we share, it is easy to ascertain that we understand each other, for example by receiving what we expect having asked “could you pass the salt please?”. We may not know the chemical formula of table salt, and we almost certainly cannot guarantee that the content of the jar is indeed NaCl in a certain percentage, but we would be surprised if we did not get what we asked for.

Between the incommunicability of those who speak different languages and everyday communication, which with good reasons we take for granted, there are multiple cases, as for example when we pick up a dictionary to learn the meaning of a word we did not know (“I have just come across the word «peritext»: what does it mean?”). This is also because we cannot always answer the questions “what is ...?” or “what does «...» mean?” by pointing to objects in front of us, for example when we are referring to entities that are not spatiotemporally located (“what is sodium chloride?”). This is why we resort to vocabularies, essentially recursive texts that use words to make the meaning of (other) words comprehensible. With the consequence that either some definitions are circular, in the sense that the definition of ‘A’ contains the term “B” and the definition of ‘B’ contains the term “A”, or some terms are used in definitions without their meaning being defined in the vocabulary itself: the former is the solution of generalist vocabularies, the latter of technical-scientific vocabularies.

The fundamental content of a vocabulary is a list of pairs

lemma: definition

(in the lexicon of terminology, terms whose meaning is to be defined are called “lemmas”), for example (from the International Vocabulary of Metrology, VIM, JCGM 200, Guide ISO/IEC 99:2007)

measurement unit: real scalar quantity, defined and adopted by convention, with which any other quantity of the same kind can be compared to express the ratio of the two quantities as a number

Hence, ternary relations are at stake here, between (i) the entities one wants to talk about, (ii) the linguistic terms one wants to talk with, and (iii) the definitions formulated to ensure that the terms refer to the entities. And indeed in linguistics these relations are sometimes presented as “triangles of meaning”, or “semiotic triangles”.

Such relations are particularly important in the technical-scientific field, where communication is expected to be unambiguous and specific, and thus such that with a certain term everyone refers to the same entity (unambiguity) and only to that entity (specificity). Put differently, a good (lemma: definition) pair provides the necessary and sufficient conditions for correct communication about the entity named “lemma”.

To achieve this, at least three kinds of problems must be solved:

1.  the entity about which one wants to communicate must be sufficiently well identified; for example, if it is plausible that everyone agrees that the metre is an example of a unit of measurement, what about the degree Celsius? and the degree on the Beaufort scale of wind force? It should be clear that an agreed answer to these questions, which is a matter of the application domain, in this case metrology, is a prerequisite for formulating a good definition;
2.  the definition must be well formulated, so as to be unambiguous and specific, and first of all comprehensible to those to whom it is addressed; for example, the definition above assumes that the meaning of the term “real scalar quantity” is known to the intended readers of the vocabulary, a condition which may not be true; one could then modify the definition, or even introduce a definition for ‘real scalar magnitude’ in the VIM;
3.  the lemma must be chosen so as to recall its meaning, and thus the entity to which it refers; for example, the lemma “measurement unit” might be considered not sufficiently specific, since it refers to entities that are used also outside measurement; on the other hand, even though the history of the use of a term may lead it to no longer correspond to its literal meaning, it might be reasonable to retain it, for reasons of habit and familiarity.

In this framework neither the definition nor the lemma are expected to be true or false. Rather, a lemma should be chosen to be comprehensible and to ensure, as far as possible, the consistency of its defined meaning with its etymon (if it is known) and its inherited meaning from the past (if it exists). During the construction of a vocabulary, it certainly makes sense to ask whether a certain lemma has been used with a certain meaning up to that point, but a positive answer does not make the corresponding definition for that lemma true, but merely consistent with current usage. That “measurement unit” has in the VIM the meaning given by the above definition is (trivially!) true, but that term has no (true) meaning in itself. Hence, truth does not apply to meanings, nor to lemmas, nor to definitions, nor to relations between lemmas and definitions.

And the same thing applies to translations. When it came to translating the third edition of the VIM into Italian, it was noted, for example, that the English term “measurement unit” should be rendered in Italian as “unità di misurazione”, but it was eventually decided to go for “unità di misura” (translatable back to English as the questionable “measure unit”). This was agreed not because considered to be “truer”, but because it was decided to be preferable to maintain a semantic imprecision (‘misurazione’ and ‘misura’ are different concepts) than to break a long lexical tradition.

These simple considerations may be of some use in helping us to clarify where the problem really lies when we become aware of any misunderstanding in technical-scientific terminology. Given this, let us propose a few examples in the context – of obvious social and not only scientific-technical importance – of the metrology in the laboratories of healthcare facilities.

In the case of the laboratories of regional and state health facilities in Italy, more than thirty thousand technicians, graduates or diploma holders, as well as several thousand graduate specialists operate: doctors, biologists and chemists, work. No data is available in this regard for private laboratories, where, moreover, these figures often also work in testing laboratories (water, food, forensics, and the like). It is therefore not surprising that the ISO 15189:2022 technical standard for the accreditation of medical laboratories prescribes in point 7.3.6, “Documentation of examination procedures”, that “... b) Procedures shall be written in a language understood by laboratory personnel ...”.

A vocabulary of metrological terms in a natural language, like English or Italian, is generally the outcome of a process that starts with guidance documents, such as the VIM, or international standards, themselves the result of scientific research, whose content is translated to be more easily accessible by manufacturers and users of diagnostic devices. For example, the five-part ISO 18113 series on information that manufacturers should provide with reagents and instruments was recently published, the first part of which, ISO 18113-1, contains a list of terms and definitions. Other standards deal explicitly with the relationships between laboratories and manufacturers of diagnostic devices, and typically contain sections of terms and definitions. In particular, ISO 17511 and ISO 21151 deal with calibration, harmonisation, and standardisation of methods, and ISO 15198 with quality control, with explicit instructions to manufacturers to provide laboratories with means and procedures.

In this situation, in which the official translation into Italian of the ISO standards is the responsibility of Ente Nazionale Italiano di Normazione (Italian National Standardization, acronym UNI), the Italian Society of Clinical Pathology and Laboratory Medicine (SIPMeL) has produced the document L5Q15, in which recommendations are proposed for the translation into Italian of the headwords appearing in the “Terms and definitions” section of the main ISO standards of metrological relevance to laboratory examinations (and therefore freely accessible through the ISO Online Browsing Platform (OBP), https://www.iso.org/obp/ui), also with reference to what is contained in the UNI CEI 70099:2008 standard, i.e., the Italian translation of the VIM.

L5Q15 notes that the terminology of technical standards influences numerous internal documents within organisations, including procedures, regulations, checklists, and therefore it must be introduced with great care. In addition, when translating (for example in Italian, but clearly not only) a technical standard one must critically manage nuances, “false friends”, homonyms, conflicts, ambiguities, useless anglicisms: a passive translation is not enough, but one must choose the most effective terms, less exposed to the risk of ambiguities, misunderstandings, or even abuse. Also for this reason, some lexical choices made in the process of translating the VIM into Italian were considered a source of ambiguity in the context of SIPMeL, and therefore the L5Q15 document made alternative proposals for them. Some significant examples of these recommendations can be found in the article in References.

The recent publication of ISO 15189:2022 has given a strong impetus to the renewal of terminology (the “Terms and definitions” section of this standard contains as many as 32 entries), and therefore SIPMeL has devoted special attention to it, noting critical points for example regarding definitions of concepts such as ‘measurand’, ‘metrological comparability’, and ‘ordinal quantity’. Other reflections, and at times perplexities, have emerged in SIPMeL from the analysis of metrologically relevant terminology in recent documents of the European Committee for Standardisation (CEN) (such as those for molecular tests, in which risks of ambiguity between terms in the clinical and metrological fields, and relating to quantitative and qualitative results are noted) and the Clinical and Laboratory Standards Institute (CLSI), the organisation representing the United States in the Technical Committee ISO/TC 212, “Clinical laboratory testing and in vitro diagnostic test systems”.

Conclusions

The terminology of metrology for medical laboratories / healthcare facilities is still a moving target, that can be improved along at least three complementary directions, by a) developing translations that are terminologically correct and at the same time effective for practitioners, b) improving the consistency of technical standards with metrology vocabularies, and c) improving metrology vocabularies themselves, to bring their contents as close as possible to users' needs. The interested scholars may find here challenging opportunities to provide their contributions to these developments.

References.

1. Mari L, Narduzzi C. Verso una nuova edizione del Vocabolario Internazionale di Metrologia (VIM)? Tutto Misure 2021;XXIII,1:81 -85.
2. Pradella M. Le parole per i laboratori nel vocabolario internazionale - Words for laboratories in the international vocabulary. La Rivista Italiana della Medicina di Laboratorio 2023 Settembre;19(3):206-11. DOI: 10.23736/S1825-859X.23.00198-6