An Historical Analysis of the Division

Concept and Algorithm Can Provide Insights

For Improved Teaching of Division
Will Windsor & George Booker
Griffith University

T he evaluation of historical literature has rarely been used to support the theory that a
conceptual understanding of mathematical processes improves learner outcomes. This
research paper has been motivated by historical developments in mathematics and their
effects on modern teaching strategies. Many of the rotely-taught procedures used in today's
classes have been developed out of the methods used from the past but these have been taken
out of their cultural contexts and taught to students without reference to their background. This
paper explores historical approaches to examine the notion that a conceptual understanding is
essential in developing an understanding division.

Introduction
All of society benefits from the technological, systemic, commercial and medical
advances created by mathematics. The simplest of mathematical ideas have proven to be
a catalyst for ingenious human achievement. Consider the notion of zero and the affects
that the full understanding this concept has had on Western civilisation. For example, the
essence of digital technology assigns the digits zero and one to electric pulses that in turn
create actions. This binary code entwines the networks of the global village. Mathematics
education should be designed and delivered to students in a manner that is beneficial to
their future requirements and fruitful participation in the global techno-democratic
communities in which they live. Otherwise, as Kline (1980, pp. 5–7) argues, there is a
travesty in the intellectual quality of mathematics caused by limitations in the current
understanding of mathematical concepts.
Solving practical problems that related to the distribution of wealth, resources and
food certainly provided an early impetus for the division concept in most societies, it
evolved out of a need to share items equally or in some specific ratio. For example, the
oldest known example of a division problem (Ifrah, 1998, pp. 121–122) originates from
Sumerian civilisation and was discovered at the Iraqi site of Fara (Šuruppak). A tablet
thought to date from around 2650 B.C. provides evidence that all the formal elements of
the division concept were used for the distribution of grain. There is a dividend, a
divisor, a quotient and a precise example of a remainder. The problem seeks to find the
division 1 152 000 "granary of barley" between a certain number of people so each

172
 An Historical Analysis of the Division Concept

person received 7 "granary of barley". It correctly states the quotient as being 164 571
with a remainder of 3.

Errors in division
Rotely acquired methods that have little mathematical relevance constitute many of the
reasons for some students' poor understanding of mathematical concepts. Evidence
(Booker et al., 2004, p. 302–331) suggests that in some classrooms, the formal teaching
of the division concept and algorithm is based on memorised rules. Procedures are
described using inappropriate language that has little relationship to mathematics. For
example the act of 'putting a number into another', 'the bringing down' of digits or the
process of divide, multiply then take-away highlight some of the inappropriate language
witnessed in classrooms. The language simply explains the recording process. Research
suggests that inappropriate language confuses learners and strongly hinders their
understanding of the division concept (Booker et al., 2004, pp. 302–331; Mulligan &
Mitchelmore, 1997; Tirosh, Graebe,r & Glover, 1990).

Figure 1.
Three common division errors exhibited by current year six students.

The first example demonstrates difficulties when zero needs to be recorded in the
algorithm. Zero poses significant problems for many students and its abstractness of the
needs to be appreciated and the concept needs to be carefully taught in a meaningful
sequence of lessons. Booker et al., (2004, p. 57) contends that zero causes many
difficulties and that the notion of writing something for nothing is rather odd for young
children. The second example exhibits difficulties with place value and renaming and
shows that a mastery of basic facts will not necessarily lead to students completing the
algorithm correctly. In the third example, the student displays poor recording strategies,
again exhibiting a lack understanding concerning the significance of place value. All three
examples exemplify the necessity of students having secure knowledge of zero, place
value and renaming before algorithms can be attempted successfully.
While explicit linking among materials, language and recording for the addition and
subtraction processes ensures that students' initial ideas fully develop, teaching sequences
for multiplication and division frequently move away from these successful strategies. In

173
Stimulating the "Action" as Participants in Participatory Research

particular, there is greater use of the more abstract symbolic representations of division
and reference to a language that has little mathematical integrity or respect of the place
value system. This example below from a widely used textbook demonstrates the heavy
emphasis placed on procedures with the sole objective of obtaining an answer.

Figure 2.
Current year 6 mathematics text (Parker et al., 2000, p. 39).

Teaching the division concept and subsequent algorithm requires educators to

initially explore with students, the relationship between the language of division
supported with materials. Only after children have had ample experiences with
describing and representing division experiences should the symbolic representations be
introduced (Kouba & Franklin, 1993). Madsen, Smith, and Lanier (1995) suggest
students' understanding of concepts is greatly enhanced by the use of problem solving,
technology and manipulative activities and need to be supported with meaningful
language. Children's intrinsic awareness of division can provide the catalysts for them to
fully comprehend the division concept. Appropriate teaching sequences focussed on the
underlying mathematical ideas can then provide the basis for developing a deep
understanding of the division process at the same time fostering improved attitudes and
greater confidence in mathematics.

Different approaches to division

The development and advancement of early mathematics was based primarily on
practical applications, driven by a need to organise, manage and document the activities
of a society. The cornerstone of all early mathematical ideas was arithmetic and
measurement (Mankiewicz, 2000, pp. 39–146). Arithmetic emerged from the contact and
dealings that people had amongst each other and the importance humans placed on
describing and recording the environment that they participated in. Initially the use of
whole numbers and the computational method of addition provided the basis and
allowed for the development of the other computational methods of subtraction,
multiplication and division (Motz & Weaver, 1993, p. 33). For example, Howard Eves
(1964, p. 29) states that early mathematics required the computation of a usable calendar,

174
 An Historical Analysis of the Division Concept

the development of a system of weights and measures to serve in the harvesting, storing
and apportioning of food, the creation of surveying methods for canal and reservoir
construction and for parcelling. Furthermore the evolution of financial and commercial
practices has been crucial in the development of mathematics as the mercantile and
trading communities of different civilisations developed mathematical ideas, processes
and recording mechanisms appropriate to the needs of their societies.
The Hindu-Arabic system of numeration and computation has flourished because of
its ability to adapt to different situations. No other number system has been able to
organise, evaluate and manipulate information as successfully (Smith, 1958, pp. 9–18).
Nonetheless, two other significant societies, Ancient Egypt and China, also developed
powerful number systems and computational procedures for division in response to their
political systems and practical economic requirements. Owing to their geographical
position, these two societies undoubtedly influenced the development of the dominant
Hindu-Arabic numeration and computation system, but there are significant differences
in the approaches they took. An examination of the understanding the three number
systems provided for the division concepts and procedures points to the somewhat
natural origin of some of the ineffective teaching strategies still used in classes today and
also highlights the meaning needed for successful division processes.
The additive nature of Egyptian numeration influenced the arithmetic methodology
of the Egyptian scholars. Multiplication was performed by a succession of doubling
operations based on the fact that any number can be represented as a sum of powers of
two (Eves ,1964, p. 39). Calinger (1999, p. 47) states that Egyptian scribes and merchants
took duplation for granted and did not question its validity. Division was overtly linked
to this multiplication process and the inverse process occurred to solve arithmetic
problems. Division in ancient Egypt might be best described as a second kind of
multiplication, where the multiplicand and the product were given to find the multiplier
(Bunt, Jones, & Bedient, 1976, p. 14). The two operations were so closely linked that it
did not require scholars to teach another process. They took the existing process and
allowed for the manipulation of data to support both processes effectively.
Gillings (1972, p. 19) articulates that Egyptian scribes considered "What must I
multiply by 8 to get 184?" rather than dividing 184 among 8. Using an additive strategy
that relied on an understanding of doubling the Egyptian division process required the
individual to
(a) Tabulate the factors and multiples of the divisor.
(b) Locate the numbers from the right hand column that would add to or as close to
the dividend, which then would be checked marked (/).
(c) Then add the multipliers corresponding the checked numbers that would give
the quotient.
If there was a remainder then fractions were introduced. It must be noted that Egyptians
only used fractions where the numerator was 1 with the exception of ⅔ and ¾ (Bunt,
Jones, & Bedient 1976, p. 14). Figure 3 highlights the four step procedure that the
Ancient Egyptians used.

175
Stimulating the "Action" as Participants in Participatory Research

184 divided among 8 = 23

1 8/
2 16 /
4 32 /
8 64 /
16 128 /
23 184
Figure 3.
The additive strategy used by the Ancient Egyptians.

The division strategies of the Egyptians met the practical needs of their society. The
distribution of bread and beer as pay in given quotas was important in the organisation
of resources. It is their practical understanding of the process that allowed the Egyptians
to accurately distribute resources in a mercantile system based on bartering. Yet, because
mathematics was based on practical notions rather than theoretical conceptualisations the
development of Egyptian mathematics remained idle. The systems and processes that
were in place solved the problems that were presented to the Egyptian society. Dunham
(1994, p. 181) argues that the scholars and merchants of Egypt may have had
cumbersome processes but invariably they always provided correct solutions that served
their society successfully for over 3000 years.
Prior to the Europeanisation of Chinese Mathematics, astronomers and scholars
documented the computational strategies used in Chinese society. These manuals
describe the strategies and procedures needed to carry out the operations of addition,
subtraction, multiplication and division. The earliest of these is the Ten Computational
Canons, a set of 12 textbooks. Under the Sui dynasty (518–617 AD) and Tang Dynasty
(618–907 AD) the textbooks were used at the School for the Sons of the State or
guozixue (Martzloff, 1988, p. 15). Written in approximately the fifth century by Sun Tsu
the Sunzi Suanjing Trilogy provides an insight into the methodology of division and
multiplication. Importantly the manuals emphasised that some of the procedures used in
China are found in other societies emphasising the interconnectedness of this
geographical area. However, there are some computational techniques that are peculiar
only to China (Martzloff, 1988, p. 217–221).
Martzloff (1988, pp. 217–218) states the Sunzi texts regarded division as the inverse
of multiplication and directly linked the algorithm to multiplication. Similar to other
societies the treatise only explained division using small numbers and in case of the Sunzi
texts the divisor is never greater than 9. The Sunzi method of division relies on an
understanding of place value. There are examples in these historical texts where the
remainder is renamed to complete the algorithm (Mikami, 1974, p. 29). With the aid of
counting rods and a checkerboard (refer to Figure 5) the division algorithm was

176
 An Historical Analysis of the Division Concept

completed using methods similar modern short division. The following example
translates how the algorithm would be presented on a checkerboard.

Problem 1 hundred 10 tens 40 ones The remainder

100 divided by divide by 6 → divided divided is recorded as
6. one is not by 6 → 1 ten by 6 → 6 ones a fractional
divisible by 6. is recorded is recorded part of the
above the in the divisor.
tens place. ones place.
Quotient 1 16 16
Remainder 100 100 40 4
Divisor 6 6 6 6
Figure 4.
Explanation of the Chinese division algorithm.

Figure 5. Figure 6.
A Chinese Master teaching the art of calculation. An example of how the rods are laid out.
(Ifrah, 1998, p. 284–285)

177
Stimulating the "Action" as Participants in Participatory Research

With the advent of the abacus in the fourteenth century, Chinese mathematics
became even more procedurally based. Using short division formulae and rotely learnt
procedures, recalled as rhymes or slogans, the Chinese could complete a myriad of
division problems. The essence of these procedures and rules ensured that computations
would be completed quickly as little or no recording was required. As practitioners
became more proficient at bead arithmetic, so did the scope for solving more advanced
computational problems. For example the fei gui jue rules (formulae for flying division)
for divisors between 11 and 99 were developed. The rules stipulated particular
procedures that needed to be memorised for every two-digit number. In the journal The
Abacus (1983, p. 44 cited in Martzloff, 1988, p. 221) the problem 4368 divided by 78 = 56
is calculated in three movements by simply applying the rules for the 78. As Martzloff
(1988, p. 220) states, the memorisation of rules and processes is borne out of practical
circumstances rather than perambulated reasoning. The Chinese view and practice of
division could also be summarised in this manner.
The millennium following the fall of the Roman Empire saw political
experimentations and religious reforms initially hinder the growth of mathematical
knowledge in Europe. Smith (1951, pp. 361–362) states that mathematics knowledge and
activity generally flourish where there is a stable political environment. During this
period the Chinese, Indians and Muslims were advancing mathematical thought
especially in the area of numeration and arithmetic. Leonardo of Pisa (1180–1250) better
known as Fibonacci, was hugely influential in transmitting the Hindu Arabic system of
numeration and computation. The son of a merchant, Leonardo of Pisa spent his
formative years travelling throughout Northern Africa. It was in Egypt, Syria and Algeria
that he came into contact with the teachings of Muslim scholars. With encouragement
and support from his father and his Muslim teachers the young Fibonacci immersed
himself in the mathematical knowledge of the area. He developed a deep understanding
of Hindu Arabic numeration and computation systems. He was rightly convinced that
the Hindu Arabic number system was far superior to the Roman numbers used in
Europe. This knowledge provided the impetus for the publication of Liber abaci or Book
of Calculation in 1202.
Due to the influences of the Hindu Arabic system it is understandable that the
primary method of recording the division process was the scratch or galley method. The
method can be traced to Eastern societies, with the Hindus employing sand tables to
record the process (Boyer, 1989, pp. 200–201). Arab writers from the time of al-
Khowarizmi (c.825)—the great mathematician and father of the algorithm—used
variations of the scratch method. Numbers were written onto the sand table and once
they were used in the algorithm were 'scratch marked' or crossed out. Maximus Planudes
(c. 1340 cited in Smith, 1953, p. 136) explained that the scratch method though difficult
to perform on paper, would naturally lend itself to the sand abacus. He continued that
because of the necessity for erasing certain numbers and the writing of other numbers in
their place, it would give rise to much confusion where ink was used, but on a sand table
it is easy to write other numbers in their place.
There are different variations of the basic scratch scheme. The algorithm presented
below is similar to the example from Tartaglia (cited in Ball, 1960, pp. 192–194). The

178
 An Historical Analysis of the Division Concept

position of the number in a column is significant not the position of the number within
the row. As the method became popular its name was Latinised to galea (galley) or
battello (boat) method. Practitioners observed that once the algorithm was completed it
seemed to leave an outline of a ship's sail as exemplified by the example below.
Interestingly at the time some teachers even encouraged their students to illustrate their
recording with the outline of a ship and its sails (Smith, 1953, p. 138; Swetz, 1987, p. 214;
Gittleman, 1974, pp. 108–110). This method was efficient and appropriate for the social
and economic requirements of the time. It demanded fewer figures than other
approaches and this was advantageous as paper and quill pens were still relatively
expensive. Mathematicians up to the sixteenth century favoured this method; indeed it
had many advocates up to the close of the eighteenth century (Smith, 1953, p. 140).

Figure 7.
Galley division 16th century (Smith, 1953, p. 138).

Figure 8.
Treviso arithmetic–galley division algorithm (Swetz, 1987, p. 95).

This algorithm demonstrates the method generally used in Europe especially in the
great Italian trading centres of the fifteen and sixteenth century. In the Venetian
mathematics textbook, Treviso Arithmetic, it shows 9065 divided by 8 is 1133. The
example in the textbook is clarified by a set of instructions that demonstrates, that in
Europe some computation processes utilised rotely memorised strategies. The
mathematician's intuitive understanding of division was challenged when problems arose
that did conform to these strategies. For example division problems where the divisor

179
Stimulating the "Action" as Participants in Participatory Research

was greater than the dividend or when new numbers evolved such as negative integers
proved problematic. Significantly due to its complexities, difficulties concerning division
were been documented. For centuries early textbooks gave only simple examples and
scholars such as Gerbert (c.980), Rollandus (1424), and Pacioli (1494) document these
difficulties. They would often state that the concept and algorithm could perplex even
the most competent of mathematician. Interestingly only simple examples with small
numbers that conformed to this limited conceptual framework were used (Smith, 1953,
p. 132). No doubt the poor understanding of the Hindu-Arabic numeration system and
its nuances hindered a deeper comprehension. Many of the generalisations were based on
understanding only the division of whole numbers.
Another issue that hindered the conceptual development of division was the
dominant view that division was repeated subtraction. Division treated as repeated
subtraction first came to prominence in Europe during the Roman Empire. Due to the
additive nature of the Roman number system repeated subtraction was easier to use to
solve division problems compared to other computational methods of the time. Suzuki
(2002, p. 275) suggests that the popularity of the repeated subtraction method was due in
part because practitioners could use mechanical devices like an abacus or counting board.
After the Hindu-Arabic system became the prevailing numeration system in Europe
repeated subtraction remained popular, as little or no understanding of the new place
value system was required. The focus of this procedure was the reliable and accurate
attainment of an answer involving natural numbers. Similarly to the Chinese, the
memorisation of the division process originates from the intuitive understanding of
division rather than a deep mathematical understanding.

Example of Repeated
Subtraction Method
1128 divide 36
360 10
768 .
360 10
408 .
360 10
48 .
36 1
12 .
Answer= 31 remainder 12
Figure 9.
Example of algorithm using repeated subtraction.
As to the first I say that division is the operation of finding, from two given numbers, a third
number, which is contained as many times in the greater third number as unity is
contained in the less number. You will find this number when you how many times the
less number is contained in the greater…of the third thing which is to be noted, that the
number which is to be divided is always greater than or equal to, the divisor. Treviso
Arithmetic (Swetz, 1987, p. 85).
180
 An Historical Analysis of the Division Concept

It is impossible to fix an exact date to the origins of long division, similar to that used
in schools today, due to the fact that it developed gradually (Smith, 1953, p. 140). The
earliest printed example appeared in Caldrini's 1491 text though it was not until the end
of the seventeenth century that long division, as we know became well established. In
Italy where the method originated it was referred to as the Danda method, which means
'by giving'. The term originates from the fact that during the division process, after each
subtraction of partial products, another figure from the dividend is 'given' to the
remainder. A significant and notable aspect of the danda method is the placing of the
quotient above the dividend as it automatically helps to locate the decimal point (Smith,
1953, p. 143). It could be suggested that the introduction of decimal notation for
fractions by Pitiscus in 1608 created the need for a method that would allow for decimal
fraction notation to be easily recorded. Importantly as decimal fractions became more
widely accepted so did the Danda method; the alignment of places ensured that the
quotient could be recorded as a decimal fraction.

Figure 10.
First printed example of the Danda Method or long division
(Caldrini, 1491 cited Smith, 1953, p. 142).

Implications for the teaching of division

Research in mathematics education clearly suggests that a conceptual understanding of
the numeration and computational aspects of this discipline improves learner outcomes
and teacher practice (Booker, 2004, pp. 10–30; Neuman, 1999, Madsen, Smith, & Lanier,
1995; DeFranco & Curico, 1997). In contrast children at a primary school level are still
being taught concepts and algorithms at a procedural and mechanical level. Further, the
strategies that some teachers use is based on rotely learnt rules and procedures that
encourage a naïve understanding of mathematical concepts (Hart, 1983).

181
Stimulating the "Action" as Participants in Participatory Research

Figure 11.
Examples from the tutorial website themathpage.com that demonstrate the use of a mechanistic
approach to the division algorithm using inappropriate language and confusing recording
strategies.

182
 An Historical Analysis of the Division Concept

This is due in part to poor teacher training and community expectations that require
children to be fluent in all computational methods at the expense of a deep and
mathematically correct understanding of computation. Twentieth century mathematics
teaching and learning is insufficient for the new cohort of students. Booker (2003) states
that students need to develop an understanding of the conceptual models that provide
meaning to the underlying concepts.

Conclusion
Evidence suggests that the Ancient Egyptians, pre-revolutionary China, and the
Europeans of the middle ages have found the concept of division difficult to fully
comprehend. For many societies, the need to avoid computational errors forced them to
adopt mechanical procedures that were effective for their communities. These
procedures were challenged as the need to understand division in alternative settings
presented problematic situations. Conceptual issues that arose outside their preconceived
frameworks of division were dismissed or not even considered as relevant. Importantly,
educators and the institutions responsible for teaching must be aware that unless a clear
concept of division is developed then children will dismiss the more abstract issues
relating to division as too difficult. Unless a clear conceptual understanding is developed
then the ability for children to adapt division to various problem situations will be limited
and the basis of their knowledge will be challenged.
Within the context of division, Hedges, Huinker, and Steinmeyer (2005) stated
unequivocally that most teachers never really understand the division concept and
conversely have difficulty explaining it to their students. No matter how well the
procedure is understood, a conceptual understanding is necessary. As students move to
the more abstract ideas of fractions, negative numbers, matrices, and algebraic equations,
the conceptual understanding is essential for ensuring that students progress from their
intuitive ideas, moving onto elementary concepts, then progressing to more advanced
mathematics. With a school system that advocates and promotes written computation,
surely it is conceivable to expect that the division concept and algorithm be presented to
students correctly and effectively. As Booker (2003) argued, an examination of the
historical paths that arose from moving from procedures to produce accurate
conceptually correct results appears to be a powerful way of providing an insight to both
students and teachers.

References
Ball, R. (1960). A short account of the history of mathematics. New York: Dover Press.
Booker, G. (2003). Multiplicative thinking: Procedural rather than conceptual knowledge as a
source of difficulties for teachers and their students. In D. Holton & I. Reilly (Eds.), Proceedings
of the Fourth Southern Hemisphere Symposium on Undergraduate Mathematics (pp. 43–49. Queenstown
New Zealand: Delta.
Booker, G., Bond, D., Briggs, J., & Davey, G. (2004). Teaching primary mathematics. Sydney:
Longman.
Boyer C. B. (1989). History of mathematics. Canada: John Wiley & Sons.
Buffum Chace, A. (1979). The rhind mathematical papyrus. Ohio: Mathematical Association of
America.

183
Stimulating the "Action" as Participants in Participatory Research

Bulgar, S. (2003). Children's sense making of division of fractions. Journal of Mathematical Behavior,
22, 319–334.
Bunt, L. N. H., Jones, P. S., & Bedient, J. D. (1976). The historical roots of elementary mathematics. New
Jersey: Prentice Hall.
Calinger, R. (1999). The contextual history of mathematics. New York: Prentice Hall.
DeFranco, T. C., & Curico, F. R. (1997). A division problem with a remainder: Embedded across
two contexts. Focus On Learning Problems in Mathematics, 19(2), 58–71.
Eves, H. (1964). An introduction to the history of mathematics. New York: Holt, Rinehart & Winston.
Gillings, R. J. (1972). Mathematics in the time of the pharaohs. New York: Dover Publication.
Gittleman. (1974). History of mathematics. Ohio. Merrill.
Gullberg, J. (1997). Mathematics: from the birth of numbers. New York: W. W. Norton & Company.
Hedges, M., Huinker, D., & Steinmeyer, M. (2005). Unpacking division to build teachers'
mathematical knowledge. Teaching Children Mathematics, 11(9), 478– 483.
Ifrah, G. (1998). The universal history of numbers–from the prehistory to the inventing of the computer. Harvill
Press: London.
Kline, M. (1980). Mathematics: The loss of certainty. New York: Oxford University Press.
Madsen, A. L., Smith, P., & Lanier, P. (1995). Does competency orientated instruction enhance
computational competence? Focus on Learning Problems in Mathematics, 17(4), 42–64.
Mankiewicz, R. (2000). The story of mathematics. London: Cassel & Company.
Martzloff, J. C. (1988). The history of Chinese mathematics. New York: Springer
Mikami, Y. (1974). The development of mathematics in China and Japan. New York: Chelsea Publication
Company.
Motz, L., & Weaver, J. H. (1993). The story of mathematic. New York: Plenum.
Mulligan, J. T., & Mitchelmore, M. C. (1997). Young children's intuitive models of multiplication
and division. Journal for Research in Mathematics Education, 28(3), 309–330.
Parker, A., McSeveny, A., Collard, R., & Thomas, A. (2000) Queensland signpost maths 6. Sydney:
Pascal Press.
Smith, D. E. (1958). History of mathematics (Vol. 2). New York: Dover Publications.
Swetz, F. J. (1987). Capitalism and arithmetic: The new math of the 15th century. Illinois: Open Court
Publishing.
Suzuki, J. (2002). A history of mathematics. New Jersey: Prentice Hall.
Tirosh, D., Graeber, A. O., & Glover R. M. (1990). Pre-service teachers' choice of operation and
division word problems. In W. Geeslin & K. Graham, K. (Eds.), Proceedings of the Sixteenth PME
Conference. Durham: University of New Hampshire, p. 57–62.

184