=Paper=
{{Paper
|id=Vol-3352/pattern1
|storemode=property
|title=An Ontology Design Pattern for Historical Metrological Practices
|pdfUrl=https://ceur-ws.org/Vol-3352/pattern1.pdf
|volume=Vol-3352
|authors=Christian Kremitzl,Christoph Schlieder,Werner Scheltjens
|dblpUrl=https://dblp.org/rec/conf/semweb/KremitzlSS22
}}
==An Ontology Design Pattern for Historical Metrological Practices==
https://ceur-ws.org/Vol-3352/pattern1.pdf
An Ontology Design Pattern
for Historical Metrological Practices
Christian Kremitzl1 , Christoph Schlieder1,* and Werner Scheltjens2
1
Cultural Informatics Research Group, University of Bamberg, Germany
2
Digital History, University of Bamberg, Germany
Abstract
The field of historical metrology studies past practices for measuring objects (or processes) as well as
the transformations of such practices. This perspective leads to research questions that are distinctively
different from those raised by the metrology for today’s natural sciences and engineering disciplines. In
this paper we explain in what way published ontologies for scientific metrology fail to capture metro-
logical practices and their transformations. We propose an ontology design pattern for modeling the
practices described in historical metrological sources. We discuss the pattern’s conceptual components
and link them to use cases from research in history.
Keywords
Ontology Design Pattern, Historical Metrology, Computational Humanities
1. Introduction
Quantitative data form the basis for most types of observations and predictions in the natural
sciences and engineering disciplines. It is therefore not surprising that the ontological modeling
of metrological knowledge, that is, knowledge about units of measurement, has been studied
since the early days of the LOD cloud [1, 2]. Today, researchers and practitioners find several
reliable solutions for modeling the quantitative aspects of their data [3]. A prominent example is
the Ontology for units of Measure (OM) [4], which has been designed from use cases in the food
industry and is currently widely adopted across a range of engineering disciplines [5]. Such
ontologies focus on use cases of unit conversion. They also address the issue of dimensional
and unit consistency. The formula 𝑓 [N] = 𝑚[kg] · 𝑎[km/s2 ], for instance, is consistent with
respect to the dimensions appearing in it, 𝑓 = 𝑚 · 𝑎, but becomes unit consistent only after
replacing km by m [4].
Scholars who study historical practices of measurement describe the specific way in which
a defined social group measures a particular quantity. They identify the group (e. g. the guild
of tailors in the medieval town of Regensburg) and the details of the measurement process
(e. g. the repeated placing of a wooden stick on a woolen cloth). Such a historical description
faces problems that are quite different from those of the measurement ontologies in engineering
WOP2022: 13th Workshop on Ontology Design and Patterns, October 23-24, 2022, Hangzhou, China
Corresponding author.
*
christoph.schlieder@uni-bamberg.de (C. Schlieder); werner.scheltjens@uni-bamberg.de (W. Scheltjens)
0000-0002-7226-8204 (C. Schlieder); 0000-0002-5209-9052 (W. Scheltjens)
© 2022 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
Workshop
Proceedings
http://ceur-ws.org
ISSN 1613-0073
CEUR Workshop Proceedings (CEUR-WS.org)
(a) Etalons (b) Transformation of measurement practices
Figure 1: (a) Physical embodiments of three base units for measuring length displayed at the town
hall of the city of Regensburg, Germany. Etalons from left to right: klafter (“klafter”), foot (“schuch”),
and ell (“öln”). [CC BY-SA 4.0 Hans Koberger, commons.wikimedia.org] (b) Schematic transformation
of two measurement practices resulting in a single unified practice (values are fictive)
[e. g. 6, 7]. In comparison with the International System of Units (SI), which we currently use
in the sciences, past systems of measurement are less complex in a number of aspects, but
considerably more complex in others. They are less complex in the sense that most past practices
of measurement refer to simple physical quantity types such as length or weight. There are very
few compound units (e. g. velocity, density), which is why the issues of dimensional and unit
consistency play a minor role. In contrast, details of the social practice of measurement turn out
to be of crucial importance. This involves, for instance, describing the delicate balance between
local measurement conditions, the legal regulation of measurement practices, and the latter’s
hesitant adoption by the community. In order to avoid confusion, we use the term scientific
metrology to denote the metrology for today’s sciences, and historical metrology for the study
of past systems of measurement [6].
When we shift the perspective from scientific to historical metrology, we move the focus of
attention from units of measurement to practices of measurement, and from the conversion of
units to the transformation of measurement practices. An example illustrates the difference.
Figure 1a shows three historical measurement standards, so-called etalons, used in the same
trading place, the city of Regensburg, Germany.
The use of more than one etalon for the single physical dimension of length points to a
phenomenon which historians have observed all over medieval and early modern Europe as
well as in other parts of the world and in different periods (see Section 2.1). Measurement
practices vary between trading places and they are often tied to specific trades or goods. The ell
from Figure 1a serves for measuring cut goods such as textile fabrics whereas the dimensions
of piece goods, such as the length of a table, are measured in feet. In many places, traders of
silk, wool and linen used different etalons, thus creating a “wool ell” distinct from the “linen ell”
and the “silk ell.” While such measurement practices may look complicated for someone used
to the SI system, they are by no means impractical. After all, in the trade of wool, there is no
urgent need for comparing the length of a wool fabric to that of a linen fabric.
Historical measurement practices are characterized by etalons which define base units. His-
torical reference works on measurement units identify a practice by the base unit and further
descriptors, such as a place descriptor. For instance, Noback and Noback (1851) identify a
measurement practice as “Elle” from “Regensburg” [8], Doursther (1840) another one as “Candi”
from “Bangalore” [9]. The extraction of knowledge graphs from such reference works provides
historical metrologists with a valuable research tool that is much easier to access than the
digital scan of the original work and that can be interlinked with other data such as historical
gazetteers.
The challenge for the ontology designer consists in modeling the measurement practices as
well as their transformations. Figure 1b illustrates what type of transformation a historian might
encounter. The measurement practices are described by conversion graphs which show units as
nodes and fictive, though plausible, conversion factors on the edges. In this simplified example,
two independent measurement practices co-exist at a first stage. Since their base units (ell, foot)
are based on different etalons, the conversion factor is determined empirically (1.942675). At a
second stage, with the goal of unifying the two practices, one unit (ell) is redefined in terms
of the other (1 ell = 2 feet). A unified practice using a single etalon (foot) results from the
transformation.
The ontology design pattern for historical metrology which we describe in the rest of this
paper is based on the ontology for the reference work [8] in the Digital Noback project.1 The
pattern addresses the core task of historical metrology, namely to describe complex measurement
practices of the past. In particular, the description includes the application domain (e. g. trade
in bulk vs. packed goods; wholesale vs. retail trade) and local conventions (e. g. allowances of
the measurement process). Historians are also interested to learn about the transformation of
measurement practices. Examples for transformations are the substitution of traditional practices
by a new legal one (e. g. the introduction of the meter), the differentiation of a practice into two
or more special purpose practices (e. g. the creation of etalons for a wool ell and linen ell), or
the simplification of a practice by redefining a unit in terms of another unit.
We describe use cases of historical metrological research in Section 2 starting with background
information on the transformation of historical measurement practices from which we obtain
a list of competency questions. We discuss related work in Section 3, followed by a detailed
description of the ontology design pattern in Section 4. We conclude in Section 5.
2. Use Cases
2.1. Transformation of Metrological Practices
Since Antiquity, humans have developed a myriad ways to measure the weight, size and
value of commodities, currencies, objects, surfaces, distances and so on. Typically, earlier
measurement procedures were valid only in restricted areas and for limited periods. This
resulted in a large geographical and temporal variation. In each country hundreds of different
units of measurement were in use, mostly with their own, locally defined values [10, 7, 6]. Since
1
https://www.uni-bamberg.de/en/digihist/projekte/digital-noback-project/
the sixteenth century, numerous accounts of measurement practices have been published in
merchant manuals, economic dictionaries, lexicons and encyclopaedia [11, 12, 13]. Prior to the
introduction of the meter around 1790, these metrological reference works fixated common
knowledge about pre-modern measurement practices. They clarified the relations and ratios
between different units of measurement, specified their use, and referenced local or regional
measurement regulations. After the introduction of the meter, metrological reference works
started to include conversions into metric equivalents (meter, liter, and kilogram) as well.
Based on a large collection of metrological reference works, Jan Gyllenbok has reproduced
hundreds of systems of units in tabular form, creating what is in fact a database of unit conver-
sions, published as a book [10]. Gyllenbok argues that systems of units are identified merely by
their type and number of base units and aims at an “easy-to-survey compilation of all known
metrology systems” [10]. As a result, he neglects the social practices of measurement in histori-
cal times. In so doing, Gyllenbok’s understanding of historical metrology differs fundamentally
from that of the Polish historian Witold Kula [6], who argued that historical metrology needs
to take into account “(. . . ) all the elements associated with measuring: systems of counting,
instruments of counting, methods of using these instruments (. . . ), the different methods of
measuring in different social situations, and finally, the entire associated complex of interlinked,
varied and often conflicting social interests.” Kula understands every measurement practice
as a social institution and thus as a means to further our understanding of “the cultural links
between nations and civilizations” [6].
Building on Kula’s seminal work, Peter Kramper has produced a detailed qualitative study of
the different stages in the long quest for unification of measures as observed and discussed in
the natural sciences, politics, and the economy [7]. Kramper has shown that, despite significant
advances in the precision of scientific measurement, earlier measurement practices persisted
in the commercial sector during much of the nineteenth century. At the same time, novel
metrological regulations set in motion processes of assimilation, simplification and unification
in metrological systems. These transformations also affected the commercial sector, causing
tension between “traditionalists” and proponents of a unified system of measurement. A decisive
step towards the latter was taken at the Meter Convention of May 20, 1875, when seventeen
countries adopted the meter as standard unit of measurement.
Metrological reference works of the eighteenth and nineteenth centuries continue to be
essential sources for the study of past practices for measuring objects (or events) as well as
the transformation of such practices. In this paper, we focus on the handbook of weights and
measures of the Nobacks [8], a “pocketbook” that consists of 1907 pages bound in two volumes.
Ever since its publication, the handbook has been famous for its exhaustiveness, reliability
and up-to-date contents. For decades to come, the “Noback” was “(...) by far the best and most
informative merchant manual in the German speaking world” [12]. The compilers of [8] aimed
at comprehensiveness. The result, which is digitally available to all, is a very dense reference
work that provides systematic accounts of the weights, measures, currency relations, exchange
rates, formal commercial institutions and informal customs of 954 places and regions around the
world — from Aachen (Germany) via Buenos Aires, Calcutta, Nairobi and St. Petersburg, all the
way to Zwolle (Netherlands). Each dictionary entry systematically describes the relations and
ratios between units as well as the different uses of “larger” and “smaller” units of weight and
measure, the persistent use in many areas of international business around 1850 of pre-modern
units of weight and measure, and their conversion into metric equivalents. In the following
sections, which tackle data modeling issues and present ontological solutions, the “Noback”
serves as an exemplar for sources of historical metrology.
2.2. Scenarios and Selected Competency Questions
We present two use case scenarios from the Digital Noback project, which illustrate the type of
research problems addressed by scholars in historical metrology. A knowledge graph extracted
from the handbook of the Nobacks [8] permits to establish how to convert between units used
in different places. In addition, and beyond what a resource such as [10] offers, it is possible to
retrieve or infer facts about the social practices of measurement. The first scenario of deprecated
measures describes this situation. Enriching the data by linking it to data from historical
gazetteers opens further avenues for research. It enables the scholar to spatially explore the data
and to search for evidence on complex transformations of measurement practices. A case in
point is the imposition of a measurement system under colonial rule as described in the second
scenario.
2.2.1. Deprecated Measures Scenario
The introduction of units of measurement by means of legislation and the adoption of the new
units by a commercial community are different processes, of which the first not necessarily
entails the second. Historians are interested in understanding what factors contribute to the
persistence of traditional measurement practices. In our use case, the scholar is studying
a geopolitical entity, e. g. the Kingdom of Bavaria around 1850, and tries to identify those
dependent entities (cities or territories) where deprecated units of measurement are still in use.
The scholar wants to learn whether there exist differences in the level of adoption between
places of trade. This transformation of old (traditional) into new (legal) measurement practices
is a typical object of study in historical metrology.
2.2.2. Colonial Measures Scenario
Measurement practices are often introduced for the purpose of fostering trade, but they also
act as a symbol of sovereignty over a geopolitical entity. Colonial rule was an extreme form of
foreign rule, in which the colonizer often imposed cultural practices such as language, religion
or, for that matter, the practices for measuring goods. Different forms of European colonial
rule on the Indian subcontinent around 1850, for instance, used different policies of imposing
measurement practices. One policy consisted in retaining the local name of a unit, e. g. to still
measure lengths in “guz,” while imposing a fixed relation to a unit of the colonizing power, e. g.
by redefining 1 guz = 32 inches. Historians study such transformations of local measurement
practices. A “simple” conversion factor provides a first clue for such a policy. Further clues are
obtained from written records of measurement practices.
2.2.3. Selected Competency Questions
Competency questions CQ 1 to CQ 5 are about the units appearing in historical measurement
practices and the conversion of these units. In many cases, historians ask a question that relates
to a particular source or compares several sources.
CQ 1. What type of quantity does the unit “guz” measure in Arungabad in 1850 according
to source S?
CQ 2. What units were used in Augsburg during the period 1820–1850 for measuring
weight according to source S?
CQ 3. How does the length unit “guz” used in Arungabad in 1850 convert to the guz used
in Delhi at the same time according to source S?
CQ 4. Which places have “Schiffspfund” as a measurement unit for weight at some time
period according to at least one source S?
CQ 5. Do the sources S1 and S2 agree on the conversion factors between the different
“guz” units used on the Indian subcontinent?
Competency questions CQ 6 to CQ 10 refer to measurement practices and their transformations.
CQ 6 for instance, asks for persisting measurement practices. Such cases could indicate problems
with adoption, i. e., an incomplete transformation of the traditional practices into the new
legal ones. CQ 8 and CQ 10 ask about a different type of transformation, the connection of a
measurement practice to another one, a step towards the unification of a system of units.
CQ 6. What were the legal practices of measuring length in Augsburg in 1820–1850
according to source S?
CQ 7. For which places in the Kingdom of Bavaria does source S report the largest (small-
est) number of traditional measurement practices?
CQ 8. For which places in the Kingdom of Bavaria does source S report practices for
measuring length that are connected by simple unit conversions?
CQ 9. Which places on the Indian subcontinent in 1850 use European measurement
practices that have been imposed by a colonial power (according to some source)?
CQ 10. Which types of measurement practices on the Indian subcontinent in 1850 have
been redefined and bound to the metrological system of a colonial power (according
to some source)?
3. Related Work
Several ontologies have been designed for publishing quantitative data within and outside the
LOD paradigm. The evaluation of metrological ontologies by Keil and Schindler [3] compares
eight such ontologies. Among these, OM 2.0 describes the largest number of units [4], whereas
QUDT has the largest number of measurement dimensions [1]. The comparison is based on an
ontology-agnostic relational model which captures common conceptualizations. This model
associates the concept of measurement unit (e. g. Newton) with the concepts of dimension
(e. g. mass · length/time2 ), decimal prefix (e. g. kilo), and system of units (e. g. MKSA — meter,
kilogram, second, ampere). This conceptualization clearly reflects a late stage in the history of
metrology when, starting in the late nineteenth century, several competing proposals were made
for universal systems of physical units. The very idea of a system of units presupposes much
more conceptual uniformity, understanding of physics and mathematical formalization than
what is encountered by historians who study the kind of measurement practices we presented
in Section 1.
On the other hand, the idea of a measurement system hides some of the complexity that
historical metrology wants to describe. This is not surprising, since the use cases of scientific
metrology as specified by [4] overlap only with respect to unit conversions with the use cases
of historical metrology which we have described in the preceding section. Historical metrology
needs a modeling that is conceptually richer. It has at least (1) to capture the grouping of units
into measurement practices, and (2) to handle the distinction between unit conversions and
the decomposition or aggregation of units outside the system of decimal prefixes. None of the
ontologies for scientific metrology evaluated by [3] serve that purpose.
We finally observe that there is a strand of research on metrological ontologies which is not
discussed in the Semantic Web literature because its axiomatization makes use of higher-order
logic [14]. Although highly relevant for inferences in the dimensional calculus, the approach, as
much as the others, focuses on scientific metrology and seems not to help with our use cases.
4. The Historical Metrology Pattern
4.1. Overview
The Historical Metrology pattern provides a solution for the problem of separately describing
measurement practices for which historical sources document independent uses (e. g. foot vs. ell).
Practice, Transformation, Unit, and Conversion are the pattern’s main components. Figure 2
shows the schema diagram of the pattern. Practice is modeled in the most detail because it is
the interaction of this component with the other three that creates the intended abstraction.
To put it in a simplified way, Practice and Transformation capture the historical aspects of
measurement, Unit and Conversion the physical aspects.
The components Decomposition and Aggregation have a supporting function. They serve to
describe the decomposition of units into subunits and the aggregation of units to superunits as
operations that are distinct from unit conversions. The schema diagram also includes compo-
nents that are left unmodeled. We adopt the graphical convention used by [15] and render those
components by dashed blue boxes. These boxes denote classes with external dependencies on
the ontology in which the pattern is used. They serve as “hooks” to a more detailed modeling.
We provide a formalization of the core components in terms of the OWL axiom patterns
described by [16]. These axiom patterns are simple in the sense that they use at most three
classes or roles, which is sufficient to express the relevant constraints of the Historical Metrology
Figure 2: Schema diagram for the Historical Metrology pattern. The yellow boxes show the classes. The
blue dashed boxes are “hooks” to potentially more complex modeling. Blue edges are object properties,
green edges are datatype properties.
pattern. For quick orientation, we add to each formula the name of the axiom pattern as specified
in the above publication. The presentation order of the axioms follows the example of [15]. That
is, axioms appear in the section which discusses the source of the arrow in the schema diagram.
Axioms using the property hasBaseUnit, for instance, are found in the Section 4.3 that covers
Practice. The OWL file of the pattern is published online in the project repository.2 The pattern
has also been submitted to the ODP portal.3
4.2. HistoricalSource
Historical research starts with the study of (mostly written) sources. It is part of the work
routine of historians to establish whether sources agree or disagree on an assertion about an
entity or a process. While sources are at the center of the methodological concerns in history,
HistoricalSource is peripheral in the Historical Metrology pattern. The concept of source is
left unmodeled because of the complexity of the interpretative process that extracts assertions
from sources. Depending on the application, the association of source, interpretation and
assertion is modeled with more or less detail. In this context, provenance can be described
by means of the EntityWithProvenance pattern from MODL [17]. The Historical Metrology
2
https://github.com/kulturinformatik/noback
3
http://ontologydesignpatterns.org/wiki/Submissions:HistoricalMetrology
pattern specifies that sources report about measurement practices, about their transformation,
as well as about conversions between units. Note that sources only report indirectly about units,
mostly through conversions that define the units in terms of some other units. Applications
which deal with sources that make assertions about units without referring to measurement
practices or conversions may add the possibility of reporting about units.
4.3. Practice
Historical practices of measurement are characterized by base units. A metrological source has
to provide at least this piece of information to be considered a witness for the measurement
practice. This is expressed by Axiom 3. In most cases, the property is also functional. While we
did not encounter such cases, there might be practices that build upon two or even more base
units. Specializations of the pattern can add a functionality axiom if needed.
Practice ⊑ ∀hasBaseUnit.Unit (scoped range) (1)
∃hasBaseUnit.Unit ⊑ Practice (scoped domain) (2)
Practice ⊑ ∃hasBaseUnit.Unit (existential) (3)
Sources refer to measurement practices by means of a Description or a combination of descrip-
tions. The historical reference works [9] and [8] use descriptions for units and for places when
they identify a practice as the one that measures in “guz” from “Bangalore.” Simple descriptions
can be modeled by data properties which specify unit names or place names. An obvious choice
for more complex spatial descriptors is GeoSPARQL, the Open Geospatial Consortium’s standard
for geospatial linked data. Some description is required to identify the practice (Axiom 6). The
Application of a practice specifies a field of application, such as measuring lengths in the trade of
cut goods or, being more specific, in the trade of woolen cloth. The component is left unmodeled.
It could refer to an ontology of historical craft and trade activities. Note that sources may be
silent about the domain. For many practices, sources attest a local Convention, e. g. allowances.
Practice ⊑ ∀describedAs.Description (scoped range) (4)
∃describedAs.Description ⊑ Practice (scoped domain) (5)
Practice ⊑ ∃describedAs.Description (existential) (6)
Practice ⊑ ∀appliesTo.Application (scoped range) (7)
∃appliesTo.Application ⊑ Practice (scoped domain) (8)
Practice ⊑ ∀followsConvention.Convention (scoped range) (9)
∃followsConvention.Convention ⊑ Practice (scoped domain) (10)
In historical measurement practices, the Decomposition of a base unit into subunits is much
less uniform and systematic than in the SI system with its decimal factors and prefixes. There is
no existential axiom since some units do not have decompositions. Aggregation is treated in an
analogous way.
Practice ⊑ ∀usesDecomposition.Decomposition (s. r.) (11)
∃usesDecomposition.Decomposition ⊑ Practice (s. d.) (12)
Practice ⊑ ∀usesAggregation.Aggregation (s. r.) (13)
∃usesAggregation.Aggregation ⊑ Practice (s. d.) (14)
4.4. Decomposition, Aggregation
Decomposition and Aggregation are mirror concepts that specify how exactly a measurement
practice decomposes a unit into subunits or aggregates the unit to a superunit. Note that
the aggregation of a base unit is not primarily a mathematical operation. It requires han-
dling a measuring device such as a yardstick. The decomposition of a base unit requires a
skilled manufacturing process which, for instance, subdivides a yardstick into 32 equal parts.
An instance of Decomposition (Aggregation) decomposes (aggregates) exactly one Unit into
decompNoOfUnits (aggregNoOfUnits) subunits (superunits). The axioms for aggregations ex-
actly mirror those for decompositions. We omit listing them as they can be found in the published
ontology.
Decomposition ⊑ ∀decomposesUnit.Unit (scoped range) (15)
∃decomposesUnit.Unit ⊑ Decomposition (scoped domain) (16)
Decomposition ⊑ ∃decomposesUnit.Unit (existential) (17)
Decomposition ⊑ ≤1 decomposesUnit.Unit (functionality) (18)
Decomposition ⊑ ∀subUnit.Unit (scoped range) (19)
∃subUnit.Unit ⊑ Decomposition (scoped domain) (20)
Decomposition ⊑ ∃subUnit.Unit (existential) (21)
Decomposition ⊑ ≤1 subUnit.Unit (functionality) (22)
4.5. Transformation
Historical metrologists are primarily interested in understanding the transformations of mea-
surement practices as well as the factors that drive them. Sources, however, often describe just
the status quo at the time of their writing. It is by combining sources from different points in
time that a picture of the transformations emerges. This dependence on the application studied
explains why it is difficult to come up with a one-size-fits-all model of transformations. The
Historical Metrology pattern refrains from specifying details beyond the fact that a transforma-
tion starts from and ends in a Practice. Information about start and end may be missing in the
sources.
Transformation ⊑ ∀fromPractice.Practice (scoped range) (23)
∃fromPractice.Practice ⊑ Transformation (scoped domain) (24)
Transformation ⊑ ∀toPractice.Practice (scoped range) (25)
∃toPractice.Practice ⊑ Transformation (scoped domain) (26)
4.6. Unit, Conversion
Historical works of reference such as [9] and [8] report conversions for many pairs of units. They
do not necessarily agree on the conversion factor or the details of more complex computations,
however. The pattern just specifies scoped range and scoped domain axioms for the interaction
of conversions and units in order to facilitate the pattern’s reuse with different ontologies for
scientific metrology.
Conversion ⊑ ∀fromUnit.Unit (scoped range) (27)
∃fromUnit.Unit ⊑ Conversion (scoped domain) (28)
Conversion ⊑ ∀toUnit.Unit (scoped range) (29)
∃toUnit.Unit ⊑ Conversion (scoped domain) (30)
Concepts closely related to Unit and Conversion are found in ontologies of scientific metrology,
for instance, the Unit concept in OM [4]. Other concepts listed in [3] are not aligned as easily.
For instance, prefixes imply the uniform decomposition and aggregation of the base unit which
historically were only introduced by the metric system, not until the French Revolution.
5. Conclusion and Future Perspectives
While several ontologies have been proposed for the field of scientific metrology, none of them
addresses the requirements of historical metrology. As we explain in Section 1 this is due to the
fact that until the nineteenth century, the practices of measuring objects in craft and trade vary
greatly between the fields of application as well as between the places where they were used.
We argue that the central concept for historical metrology is the measurement practice together
with the concept of transformation of such practices over time. In support of this view, we
describe two use cases from the Digital Noback project as well as a set of competency questions,
which historians who study measurement practices intend to address (Section 2). The Historical
Metrology pattern is extracted from the ontology created for that project. The pattern allows to
express how measurement practices interact with their transformations and how they relate to
units of measurement and their conversions. The latter two concepts act as a conceptual bridge
to the ontologies for scientific metrology (Section 4).
We formulated the constraints on the pattern in form of OWL axioms, taking care to be as
least restrictive as possible. We paid particular attention to incomplete historical knowledge.
So it is possible to specify details of a practice without knowing much about how it was
transformed. This permissive handling of the constraints should facilitate the reuse of the
pattern in existing domain ontologies. Instantiations of the pattern can easily add further
constraints if the domain under study warrants it. On the other hand, the pattern underpins
important conceptual distinctions. The decomposition of a unit and the aggregation of a unit
are treated as operations that are clearly distinct from the conversion between units.
In our future work we plan to optimize the knowledge extraction process from metrological
reference works. Some of these works have been published in revised editions for several decades,
thereby providing valuable information about changing measurment practices. Extracting and
comparing knowledge graphs from different editions could help gaining a better understanding
of the transformations of measurement practices. This would probably have an impact on the
modeling of transformations as well.
References
[1] R. Raskin, M. Pan, Semantic web for earth and environmental terminology, in: Workshop
on Semantic Web Technologies for Searching and Retrieving Scientific Data, 2003.
[2] J. Madin, S. Bowers, M. Schildhauer, S. Krivov, D. Pennington, F. Villa, An ontology for
describing and synthesizing ecological observation data, Ecological informatics 2 (2007)
279–296.
[3] J. M. Keil, S. Schindler, Comparison and evaluation of ontologies for units of measurement,
Semantic Web 10 (2019) 33–51.
[4] H. Rijgersberg, M. Van Assem, J. Top, Ontology of units of measure and related concepts,
Semantic Web 4 (2013) 3–13.
[5] F. Martín-Recuerda, D. Walther, S. Eisinger, G. Moore, P. Andersen, P.-O. Opdahl, L. Hella,
Revisiting ontologies of units of measure for harmonising quantity values–a use case, in:
International Semantic Web Conference, Springer, 2020, pp. 551–567.
[6] W. Kula, Measures and Men, Princeton University Press, 2014.
[7] P. Kramper, The battle of the standards: Messen, Zählen und Wiegen in Westeuropa
1660-1914, Walter de Gruyter, 2019.
[8] C. Noback, F. Noback, Vollständiges Taschenbuch der Münz-, Maass- und Gewichtsver-
hältnisse, der Staatspapiere, des Wechsel- und Bankwesens und der Usanzen aller Länder
und Handelsplätze, F. A. Brockhaus, 1851.
[9] H. Doursther, Dictionnaire universel des poids et mesures anciens et modernes, contentant
des tables des monnaies de tous les pays, 1840.
[10] J. Gyllenbok, Encyclopaedia of historical metrology, weights, and measures, Springer, 2018.
[11] D. Besomi, Crises and Cycles in Economic Dictionaries and Encyclopaedias, Routledge
New York, 2012.
[12] M. A. Denzel, J. C. Hocquet, H. Witthöft, Kaufmannsbücher und Handelspraktiken vom
Spätmittelalter bis zum beginnenden 20. Jahrhundert, Franz Steiner Verlag, 2002.
[13] J. Hoock, P. Jeannin, W. Kaiser (Eds.), Ars mercatoria : Handbücher und Traktate für den
Gebrauch des Kaufmanns, 1470-1820, Schöningh, Paderborn, 1991.
[14] M. Grüninger, B. Aameri, C. Chui, T. Hahmann, Y. Ru, Foundational ontologies for units
of measure., in: FOIS, 2018, pp. 211–224.
[15] C. Shimizu, R. Zhu, M. Schildhauer, K. Janowicz, P. Hitzler, A pattern for modeling causal
relations between events, in: Proceedings of the 12th Workshop on Ontology Design and
Patterns (WOP 2021), 2021, pp. 38–50.
[16] A. Eberhart, C. Shimizu, S. Chowdhury, M. Sarker, P. Hitzler, et al., Expressibility of owl
axioms with patterns, in: European Semantic Web Conference, Springer, 2021, pp. 230–245.
[17] C. Shimizu, Q. Hirt, P. Hitzler, Modl: a modular ontology design library, arXiv preprint
arXiv:1904.05405 (2019).