A Statue S Weight Problem Reedition of P

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

JULIA LOUGOVAYA

A S TATUE ’ S W EIGHT P ROBLEM . R EEDITION OF P.R AIN .U NTERRICHT 179

aus: Zeitschrift für Papyrologie und Epigraphik 220 (2021) 254–258

© Dr. Rudolf Habelt GmbH, Bonn


254

A S TAT U E’S WEIGH T P ROBLEM. R EEDI T ION OF P. R A I N.U N T ER R ICH T 179

Among the school texts in the Vienna collection is an interesting arithmetical problem, P.Rain.Unter-
richt 179 (TM 63194; 1st c., Soknopaiou Nesos), inscribed on the recto of P.Vindob. G26011e.1 The verso,
which was edited already in 1902 by Carl Wessely, preserves a list of syllables.2 The fact that the recto was
not published until 80 years after the verso probably has to do with the state of the fragment: not only is the
writing extremely faint, the papyrus has holes and seems to have been reused, with traces of washed-off ink
further obscuring the writing. H. Harrauer and P. Sijpesteijn, the editors of P.Rain.Unterricht, managed to
transcribe most of it, but the resulting text was not satisfactory and the math did not work. Upon examina-
tion of the papyrus I was able to correct some of the readings, on the basis of which all computations can
be accounted for and the overall structure of the problem restored.3
The problem in the papyrus belongs to a type of partition with a remainder that is well represented
among the mathematical epigrams in Book XIV of the Palatine Anthology: these problems give the parts of
an unknown number and the remainder left after the given parts are subtracted from the whole.4 AP XIV 2
is a good example:
εἰς ἄγαλμα Παλλάδος
Παλλὰς ἐγὼ χρυσῆ σφυρήλατος· αὐτὰρ ὁ χρυσός
αἰζηῶν πέλεται δῶρον ἀοιδοπόλων.
ἥμισυ μὲν χρυσοῖο Χαρίσιος, ὀγδοάτην δέ
Θέσπις καὶ δεκάτην μοῖραν ἔδωκε Σόλων,
5 αὐτὰρ ἐεικοστὴν Θεμίσων· τὰ δὲ λοιπὰ τάλαντα
ἐννέα καὶ τέχνη δῶρον Ἀριστοδίκου.

On a statue of Pallas
I am Pallas, wrought of beaten gold. Lo, the gold is the gift of mighty poets: Charisios gave
half the gold, Thespis one-eighth and Solon one-tenth part, then Themison one-twentieth. The
remaining nine talents and the craftsmanship were the gift of Aristodikos.
To solve the problem, that is to find the weight of the statue, one needs to find a number which can be par-
titioned into whole numbers equaling its half, one-eighth, one-tenth and one-twentieth, with the remaining
number of parts corresponding to the nine talents of gold given by Aristodikos. The number satisfying
these conditions is 40, because 40 can be partitioned into its half (20), one-eighth (5), one-tenth (4) and
one-twentieth (2) parts; subtracting the sum of these parts (20 + 5 + 4 + 2 = 31) from 40 yields the 9 parts
comprised by the nine talents given by Aristodikos. It follows that the whole, which comprises 40 parts, is
40 talents.5
1 I am grateful to Bernhard Palme and Claudia Kreuzsaler of the Papyrussammlung und Papyrusmuseum at the Öster-
reichische Nationalbibliothek for facilitating the study of the piece, to Andrea Jördens for valuable comments and, above all,
Rodney Ast for helping decipher this difficult papyrus. For support of my research, I thank the Collaborative Research Centre
933 – Material Text Cultures. Materiality and Presence of Writing in Non-Typographic Societies (Subproject A09: Writing
on Ostraca). CRC 933 is funded by the German Research Foundation (DFG) under Project Number 178035969 - SFB933. All
dates are CE, and the translations are mine.
2 Cf. Stud.Pal. II p. XLV–XLIX no. 2–5; the verso was reedited in P.Rain.Unterricht 8.
3 A photo of the papyrus is available at http://data.onb.ac.at/rec/RZ00003998.
4 For the typology of the mathematical epigrams in the Anthology, cf. P. Tannery, Sur les épigrammes arithmétiques de
l’Anthologie Palatine, REG 7 (1894), p. 59–62; F. Acerbi, Byzantine Rechenbücher. An Overview with an Edition of Anonymi
J and L*, Jahrbuch der Österreichischen Byzantinistik 96 (2019), p. 1–57, at p. 7–8, esp. fn. 34.
5 This reconstruction is based on the solution given in the upper margin of the page of Paris. suppl. gr. 384, on which the
epigram is written; for the edition of the scholion, see P. Tannery, Diophanti Alexandrini opera omnia cum graecis commenta-
riis, Vol. II, Leipzig 1895, p. 44. The image of the page is accessible at https://gallica.bnf.fr/ark:/12148/btv1b8470199g/f8.item.
A Statue’s Weight Problem. Reedition of P.Rain.Unterricht 179 255

The problem in the Vienna papyrus is remarkably similar.6 Although not in verse, the formulation has a
fable-like narrative. As in the epigram, the statue speaks in the first person and the task, which is to find the
weight of the original whole statue in mnas, is not explicitly stated, but is implied. Unlike in the epigram,
the problem on the papyrus is followed by a solution introduced by the formulaic expression “do it this
way” or “this is how to do it” (ὡς δεῖ ποιῆσαι).7 Its first step consists in adding up the parts that the statue
had lost, i.e. 1/3 + 1/4 + 1/5 + 1/6, with the sum expressed as a fractional number, 1/2 1/4 1/5. The papyrus does
not record how the sum was computed, but one can imagine two possibilities. First, one might have added
all the fractions by finding a number into which 3, 4, 5, and 6 parts can be partitioned, or resolved, that is
a number, the stated parts of which are whole numbers (we would call it the lowest common denominator).
This number is 60, with the sum of parts being 57 (we would call this the nominator of the fraction 60 57, but
__
since the ancients used only unit fractions to express fractional numbers, 57 parts of 60 was viewed as a
division or a ratio, not a fractional number). 57÷60 could then be easily partitioned into 1/2 1/4 1/5 (in modern
notation 30+ + __ = 1_ + 1_ + 1_).8 Alternatively, and perhaps likelier, the person simply knew that 1/3 + 1/6 = 1/2
____1__5__12
2 4 5
60
and used this substitution to express 1/3 1/4 1/5 1/6 as 1/2 1/4 1/5.9
The lowest number into which the fractional number 1/2 1/4 1/5 can be resolved, i.e. the number that can
be partitioned into whole numbers corresponding to its half, quarter, and one-fifth is 20, with the sum of
the parts amounting to 19 (because 1/2 1/4 1/5 of 20 is 10 + 5 + 4 = 19). The statue thus lost 19 parts, with the
remaining 120 mnas comprising one remaining part. Consequently, the weight of the whole statue is the
product of 120 mnas multiplied by 20 = 2,400 mnas. After this result is recorded, there follows a demon-
stration, ἀπόδιξις (l. ἀπόδειξις), that the obtained result is correct, in which each loss should have been
computed (i.e. the 1/3, 1/4, 1/5, and 1/6 of the whole) and then summed up. Although the papyrus is very dam-
aged in line 8, it is virtually certain that the values for each loss (i.e. 800, 600, 480 and 400 mnas, respec-
tively) were not inscribed. Only their sum was, which amounts to 2,280. Because subtracting 2,280 from
2,400 yields 120 mnas, which is what is said to be left of the statue, the problem has been solved correctly.
While the computations in the problem are clear, the curtness of style, which is characteristic of prob-
lems on papyri in general and of this one in particular, makes it difficult to reconstruct some details and
expressions. Thus, after the three elements responsible for the destruction of the statue – the sun, the winds,
and the dew – one expects a fourth of the same order, but instead there comes ὁ δὲ κατασκευάσας, “the
one who has built (it)”, or simply “furnished”. Does this stand for damage caused by a human as opposed to
the natural elements? Or did the participle lose its subject and/or direct object, which could have explained
6 The similarity of this problem to those in the Anthologia Palatina is discussed by J. Teichmann, Das Rechenbuch des
Metrodor, Hermes 148 (2020), p. 86–118, at p. 99–100, who notes that the text of the papyrus is difficult to understand, but does
not question the edition.
7 For the use of the expression ὡς δεῖ ποιῆσαι to introduce the solution, cf. for example, P.Rain.Unterricht 178 (TM 60599;
1st–2nd c., Soknopaiou Nesos) or P.Bagnall 35 (TM 65836; 2nd c., prov. unknown). Mathematical epigrams collected in the
Anthology never include solutions; the scholia in Paris. suppl. gr. 384, however, have solutions for and sometimes discussions
of most of the problems, cf. the edition of Tannery cited in fn. 5, p. 43–72.
8 Cf., for example, the solution to problem 19 in P.Cair.Cat. 10758, also known as the Akhmim mathematical papyrus
(TM 64999; late 4th or 5th c. [for this date, see P.Math. p. 3, fn. 8], Panopolis), with the discussion of it in P.Math., p. 50–51. This
problem asks for a fractional number comprised of three fractions, 1/55 1/56 1/70, to be converted into four fractions. The first
step of the solution consists of adding up the given fractions, which results in 155 parts of 3080, i.e. 155÷3080; this division is
later converted to the lowest terms, 31÷616, and then partitioned into 1/63 1/77 1/88 1/99.
9 Solutions to numerous problems in P.Cair.Cat. 10758 feature substitutions, which might be drawn from the division
table preserved in the same codex or from memory; cf., e.g., a step in the solution of problem 35, ἐν ποίᾳ ψ[ή]|φῳ 𝈪 dʹʹ; τῶν Γ
τὸ d ʹʹ, “In which calculation 1/2 1/4? In the 4th part of 3”. Although the language is the same as in the similar step in problem 19
discussed in the previous footnote, one can presume that the conversion was done off-hand. It is of course not always possible to
determine which conversions a person performing computations with fractions actually computed and which he or she arrived
at by consulting the tables or by using obvious or previously learned “identities”. For the use of division tables and computa-
tional procedures in operations with fractions, cf. W. Knorr, Techniques of Fractions in Ancient Egypt and Greece, Historia
Mathematica 9 (1982), p. 133–171, esp. p. 148–151; D. Fowler, The Mathematics of Plato’s Academy, 2nd ed., Oxford 1999,
p. 235; A. Jones, Mathematics, Science, and Medicine in the Papyri, in R. S. Bagnall, (ed.), The Oxford Handbook of Papyrol-
ogy, New York 2009, p. 338–357, at p. 340–341; J. Lougovaya, Computational Workout: Division Tables as Training Exercises,
in M. Capasso, P. Davoli, and N. Pellé (eds), Proceedings of the XXIX International Congress of Papyrology, forthcoming.
256 J. Lougovaya

by whom – or how – the final damage was caused? In addition to such ambiguity, the end of the problem
cannot be restored with confidence because of damage to the papyrus, although it is clear that some steps
in the computations were never recorded. It is likely that the text as we have it was written as notes and was
meant to be further explicated orally.
The edition below indicates only substantial changes to the editio princeps. It does not record diffe-
rences in uncertain readings, i.e. those signaled by underdots.
ἀνδριάς τις ἐνέτυχεν τῶι Δι⟦ω⟧εί· ὁ ἥλιός μου
εἴληφεν τὸ γʹ μέρος, οἱ δ’ ἄνεμοι τὸ d, ἡ δὲ δρόσος
τὸ εʹ, ὁ δὲ κατασκευάσας τὸ ϛʹ. λ⟦ι⟧⸌εί⸍πωμαι δὲ τῶν
4 μναειαίων ρ͞κ. ὡς δεῖ ποι∞σαι· σύνθες τὸ γʹ
καὶ τὸ d καὶ τὸ εʹ καὶ τὸ ϛʹ, (γίνεται) 𝈪 d εʹ εἰς κ
(γίνεται) ι͞ θ̣ καὶ τὰς ρ͞κ εἰς κ ̅ (γίνεται) xΒυ. ⸌ ⸍̣ ν ἄρα ὁ ἀνδρ[ιὰς?]
μναειαίων xΒυ. ἀπόδιξις. λαβὲ τῶν xΒυ
8 [τ]ὸ γʹ τὸ d ̣ εʹ ϛ̣ʹ (γίνεται) xΒσπ, λοιπαὶ ̣ ̣ ουσ ̣ ̣ ρ͞κ
δ ̣ ̣ κ̣ ται
1 ἐνέτυχεν σ(χοινίων) ωπδ ιβ ὧν ed.pr.; l. Διί 1–2 μετ|είληφεν ed.pr. 2 ἄνεμοι τὸ τʹ ed.pr.
3 δ’ αἱ νωμαὶ δὲ περ[ί] ed.pr.; l. λείπομαι 4 συ⟨ν⟩δίθει (l. συντίθει) ed.pr. 5 (γίνεται) σ(χοινία) 'Dχνε
εἰς κ ed.pr. 6 ποιεῖ καί ed.pr.; (γίνεται) σ(χοινία) 'Βυ ⸌ὧ⸍ν ed.pr. 6–7 ἀνδρι|ὰς λείπεται 'Βυ σ(χοινίων).
ἀπόδιξις (l. ἀπόδειξις) λοιπῶν 'Βυ σ(χοινίων) ed.pr. 8 [τ]ὸ γʹ ἐπὶ τὸ τʹ, (γίνεται) σ(χοινία) 'Βσν λοιπαὶ
ὀλίγους ρκ ed.pr. 9δ α̣ κται ed.pr.

Some statue complained to Zeus: “The sun took 1/3 part of me, the winds 1/4, the dew 1/5, and
the ‘builder’ (?) 1/6. I am left with 120 mnas.” Do it this way: add 1/3 and 1/4 and 1/5 and 1/6,
equals 1/2 1/4 1/5; resolve to 20, equals 19; and the 120 (sc. mnas) multiply by 20, equals 2,400.
Thus, the statue was 2,400 mnas. Demonstration: take from 2,400 1/3 1/4 1/5 1/6, total 2,280, the
remainder … 120 ….

1 ἐνέτυχεν τῶι Δι⟦ω⟧εί: the verb followed by the dative of the addressee often means “to converse
with”, but also “to complain to”. A fitting parallel is furnished by the opening of an Aesopic fable on
The Trampled-on Snake, ὄφις ὑπὸ πολλῶν πατούμενος ἀνθρώπων τῷ Διὶ ἐνετύγχανεν, “The snake
was complaining to Zeus about being trampled on by many people” (Fab. 213, Aesop, Corpus fabu-
larum Aesopicarum, ed. A. Hausrath, H. Hunger, vol. I.2, Leipzig 1959, p. 34). The reading of the
article τῶι is clear, but the following letters are somewhat more difficult as there seems to be traces of
erased letters; it looks like τῶι ἰδίῳ may have been written first. Alternatively, the traces could belong
to an earlier text that had not been completely washed off.
In the ed.pr. the letters following ἐνέτυχεν were interpreted as σ(χοινίων) ωπδ ιβ, “884 1/12
schoinia.” What might look like a sigma followed by an abbreviation mark is the left-hand side of
a tau, which often protrudes quite far to the left, cf., e.g., τό in the beginning of line 3. The word
σχοινίον, which is a unit of length, does not occur in the papyrus.
– μου: Harrauer and Sijpesteijn read μετ|είληφεν. Although sometimes tau has a split upper part, which
makes it resemble an upsilon, cf. the tau of τάς in line 6, it never protrudes so far up. The letter is rath-
er upsilon and the half circle before it is the left side of an omicron. For the expression λαμβάνειν +
μου, “to take away from me”, cf., for example, εἰ πολλά μου λάβοιεν ἐνέχυρα, “had they taken a lot
from me as security” in Dem. 47.74.
2 τὸ d: the symbol for 1/4 here and in lines 5 and 8 is in the form of a Latin-looking d, which is common.
It has a small horizontal stroke that intersects with the vertical just above the loop, which must have
led the editors to interpret it as τʹ inscribed mistakenly for δʹ; E. M. Bruins, who provided a mathe-
matical commentary on the problem in the ed.pr., assumed that τʹ was written for τέταρτον.
A Statue’s Weight Problem. Reedition of P.Rain.Unterricht 179 257

3 λ⟦ι⟧⸌εί⸍πωμαι (l. λείπομαι) δὲ τῶν: the beginning of the verb is messy; the writer seems to have
written the first syllable with an iota, but then corrected it to ει in superscript. The end of the line is
virtually gone, but the traces are compatible with τῶν, that is in agreement with μναειαίων in the
following line. The expression might not be conventional, but its meaning is clear.
4 σύνθες: the ed.pr. reads συ⟨ν⟩δίθει, but what looks like δί is, I believe, a wide-shaped nu with a little
inward looking hook at the bottom of the left-hand vertical (what looks like a larger hook belongs to
the upsilon); the first nu of ἐνέτυχεν in line 1 has a similar shape. The last letter of the word is cer-
tainly sigma.
5 (γίνεται) 𝈪 d εʹ: 1/2 1/4 1/5 is the sum of the parts destroyed by various agents. Strictly speaking, solv-
ing the problem does not require converting the set of unit fractions given in its statement, that is
1/3 1/4 1/5 1/6, to the set of unit fractions corresponding to their sum, that is 1/2 1/4 1/5, but one can imagine
several reasons why this was done. The original sequence was viewed as the enumeration of the giv-
ens, not as a sum, but it was the sum of what was lost that the writer wanted to compute in order to
determine what part of the statue was left. Converting the original sequence into 1/2 1/4 1/5 also meant
that further computation could be carried out with 20 parts as opposed to 60 (see introduction above),
which may have seemed easier to do. Finally, the simplification was obvious because the person
charged with solving such problems would most likely have known off-hand that 1/3 + 1/6 equals 1/2.
5–6 εἰς κ | (γίνεται) ι͞ θ:̣ the number 1/2 1/4 1/5 is resolved (cf. ἀνάλυσον εἰς for the same operation in
P.Mich. III 145 col. v.2, col. vi.6, 10 [TM 63556; 2nd c., prov. unknown]; SB XVI 12680.8 [TM 63476;
138–161, Theadelphia]) to 20, resulting in 19, because the sum of one-half (10), one-fourth (5), and
one-fifth (4) of 20 is 19. It follows that 19 of 20 parts of the statue had been lost.
6 καὶ τὰς ρ κ εἰς κ ̅ (γίνεται) xΒυ: this was read correctly in the ed.pr., but because the size of the statue
was presumed to be stated in line 1 and the numerals in line 6 were interpreted wrongly, the right
solution was not pursued, cf. comments on p. 173 of the edition. Since 120 mnas constitute 1 part of
20 (cf. ln.5–6), the whole statue, i.e. 20 parts, is 120 × 20 = 2,400 mnas.
– ⸌ ⸍̣ ν ἄρα ὁ ἀνδρ[ιάς?]: the ed.pr. has ⸌ὧ⸍ν, and omega might well be right; it can, however, also be an
eta, in which case it is the verb ⸌ἦ⸍ν. Of the last word printed here, the delta is clear; it is followed by
a long vertical, likely from rho, and then by some indistinct traces of ink. Whether the latter belong to
the word ἀνδριάς, to earlier writing or to later smudges is not clear. If the words were written in full
and not abbreviated, the last three letters would need to be placed very tightly together. The editors
read ἀνδρι|άς, but I do not see ας in the beginning of the following line.
7 μναειαίων xΒυ: although much is lost, reading μναειαίων is more compatible with the traces on the
papyrus than |ὰς λείπεται 'Βυ σ(χοινίων) printed in ed.pr. There seems to be ink to the upper right of
the upsilon of xΒυ. It might belong to the horizontal stroke marking the numeral or be a punctuation
mark signaling the end of the problem. I doubt it is an abbreviation for mnas, since the word is always
spelled out in the papyrus.
– ἀπόδιξις, l. ἀπόδειξις: “demonstration”, a common term introducing a verification that the obtained
result satisfies the conditions set out in the statement of the problem cf., for example, P.Mich. III 145
col. ii.6, col. vi.7, 11, P.Math. Dv5, Fr2, Ov7 (TM 92734; 350–375, Oxyrhynchos?).
– xΒυ: following this number are slight traces, but they seem not to belong to this text.
8 [τ]ὸ γʹ τὸ d ̣ εʹ ϛ̣ʹ (γίνεται) xΒσπ: the editors print [τ]ὸ γʹ ἐπὶ τὸ τʹ, (γίνεται) σ(χοινία) 'Βσν. The
first two letters of the line are very clear, [τ]ὸ γʹ, and the numeral xΒσπ is certain, but the writing in
between is difficult to discern. It is possible that the fractions εʹ ϛ̣ʹ were preceded by the article writ-
ten in an abbreviated form, with the omicron raised. It is virtually certain, however, that the weight of
each loss, that is 800, 600, 480, and 400 mnas, was not recorded on the papyrus, but only their sum
was given. One wonders whether the writer took a shortcut and computed the lost weight by multi-
plying 120 mnas by 19, the total number of the lost parts.
258 J. Lougovaya

– λοιπαὶ ̣ ̣ ουσ ̣ ̣ ρ͞κ: the letters are suggestive of λοιπαὶ ἐνοῦσαι, “remaining (sc. mnas) present, 120”.
λοιπαί is a natural term to express the difference, i.e. the result of subtracting the lost 2,280 mnas
from the original 2,400 mnas of the statue’s weight.
9 δ ̣ ̣ κ̣ ται: the editors read δ ̣ α ̣ κται. If these letters are indeed on the papyrus, perhaps δ°δεικται,
“has been demonstrated” or “proven”, could be read since ει can look similar to alpha. I can, however,
hardly discern any certain letters.

Julia Lougovaya, Universität Heidelberg


lougovaya@uni-heidelberg.de

You might also like