Summarizing Complex Units

By Miller Prosser, February 2018

This article demonstrates how to use built-it OCHRE features to summarize the values of complex measures, such as those attested in the cuneiform world.

A cuneiform text recording 2,793 dry liters of barley may record that number as something like: 15 GUR 2 PI 3 BÁN 3 SÌLA ŠE.BAR. This logographic rendering indicates:

15 kurru-measures + 2 pānu-measure + 3 sūtu-measures + 3 qû-measures.

The relationships of these various measures in the kurru system of measures measurements can be expressed as follows:

1 kurru (GUR)= 5 pānu (PI)= 30 sūtu (BÁN)= 180 qû (SÌLA)= 1800 aklu (NINDA) ≈ 180 dry liters

or

1 kurru = 180 liters

1 pānu = 36 liters (i.e. 180 divided by 5)

1 sūtu = 6 liters (i.e. 180 divided by 30)

1 qû = 1 liter (i.e. 180 divided by 180)

1 aklu = .1 liters (i.e. 180 divided by 1800)

So, in the example above, the values translate into dry liters as follows:

15 GUR x 180 = 2700 liters

2 PI x 36 = 72 liters

3 BAN x 6 = 18 liters

3 SILA x1 = 3 liters

Grand Total: 2,793 dry liters.



OCHRE can perform this calculation and any other type of conversion wherein the relationship between the various measures can be defined.

Define the Conversion Factor

kurru system


Consult with the OCHRE Data Service to define the various measures and their relationships.

Assign Conversion Factors to the Dictionary

Using a predefined variable called Measure, assign each of these measures to its corresponding word in the dictionary. Here, the kurru is the same kurru above from the Concepts category.

Dictionary measure

Now, when a word in a text is linked to this word in the dictionary, OCHRE understands that the word is a measure in the kurru-measure system. Therefore, OCHRE knows its value in dry liters.

The Measurement Variable

The Measurement variable is a derived variable that will look for both a number and a measure as child items among the descendant discourse units of the phrase or clause to which it is applied. The number item (a discourse unit with Type=number) and the measure item (a discourse unit with Type=word, or blank Type, which links to a Dictionary entry that is, by definition, a Measure) can be subordinate at any descendant level (that is, they do not need to be immediate children) and can be found in any order (that is, the number does not need to appear first). It is up to the project to ensure that higher-level discourse units are constructed in such a way that measures (the number and its related measure) are unambiguous within the phrase or clause tagged by the derived variable.

In this case, the parent discourse unit ("15 kurru") has been tagged with the Measurement variable. This variable looks at the child discourse units and finds "15" (the number) and "kurru" (the measure). The derived variable follows the link from the measure item ("kurru") to the Dictionary where the linked dictionary lemma is a Measure, by definition (that is, on its Properties). The Measure property identifies the linked Concept that provides the Conversion factor. OCHRE multiplies the Conversion factor of "kurru" by the number (i.e. 15 x 180) and provides the result: 2,700.

Aggregation

With each of the four discourse units above having had their respective measures calculated by the Measurement (l) property, we can now provide an aggregation clause even further up the hierarchy of discourse units. In this case, the project has defined a Variable called Volumetric summary (l) -- a derived variable of Type Aggregation -- which aggregates all occurrences of the Measurement (l) property contained below it in the discourse hierarchy.