The question as to which came first, the numerals
or the letters, was answered convincingly a century ago with the conclusion that
the “use of visible signs to represent numbers and aid reckoning is not only
older than writing, but older than the development of numerical language on the
denary (decimal) system.”
Two additional conclusions may be considered:
1- A quinary system is older than a decimal system if proven to be a subsystem of the
2- A numeral
alphabets preceded language alphabets, and indeed some numeral symbols may have
been used to construct language scripts with or without prior knowledge of their
original function. The consensus of researchers and historians of numbers is
that among primitive tribes hands and fingers have been employed for the most
common purposes of calculation since ancient times, as well as among the more
cultivated nations to the present day.
The “hawa’i” calculus is known as the “calculus of the hand” and the
numeral pictograms are known as the “hawa’i numerals”. Every concept related to
this branch of calculation is given a distinctly Arabic name, and Jahiliah
poetry and tradition attest to its ancient Arabian roots. Yet, we cannot exclude
the possibility that unspecified elements of this calculus were known to the old
South Arabians and others. One of the reasons is that both the South Arabian and
Palmyrene numeral systems, like the Arabic system, appear to have been quinary
at one point in their development, a characteristic shared by the three systems
to the exclusion of all other numeric systems for which we have records of usage
in the Middle East. In describing the numeral system as ‘Arabic’, we are not
claiming it on behalf of Northern Arabs whose language is Arabic, but simply
describing a universal system known by this name and, in the process, aiming to
From the above passage, we may conclude that the simplicity of this
finger and hand-based calculus made it possible for a relatively large number of
ordinary people to use it. The numeral shapes are permanently present on their
fingers and hands. There was no need to inscribe them lest they are forgotten.
They are there on every hand. Not every letter can resemble a numeral shape. But
creating literary alphabets came much later than creating a numeral alphabet,
and it would be natural to think of the available numeral pictograms as one of
the sources for the alphabets created at a later stage. The people of the
Maldives use the ten eastern Arabic numerals as letters in their unique
alphabet. Should we dismiss other nations from having done so in the past?
In The Story of the Alphabet, Otto F. Ege told us that our
transition from barbarism to civilisation could be attributed to the alphabet.
Those great prehistoric discoveries and inventions such as the making of fire,
the use of tools, the wheel and the axle, and even our modern marvelous
applications of steam and electricity pale into insignificance when compared to
the power of the alphabet. “Simpler as it now appears after the accustomed use
of ages, it can be accounted not only the most difficult, but also the most
fruitful of all the achievements of the human intellect.”
Very few are aware that there are seven thousand living languages in
the world today, and it would be unrealistic to believe that only a few existed
in the cradle of civilisation. Indeed, a certain unknown and crude alphabet may
have predated all recorded languages. Ego’s brilliant summary of the story of
civilisation through its alphabet suggests four stages in the development of
scripts. The first involves “picture writing” by means of a picture or character
suggesting a ‘thing’ or an ‘incident’. The next is a picture or character
“symbolising” the ‘thing’ or ‘idea’. The third is to provide a picture or
character representing the sound of the ‘thing’ or ‘idea’, and the fourth stage
is the compilation of signs suggesting the various sounds of the alphabet.
If we imagine the Arabic numerals as the alphabet of a language, we
will find that the first stage, ‘picture writing’ by means of a picture or
character suggesting the ‘thing’ or an ‘incident’, was the last stage. Moreover,
even this elementary first stage was not in fact completed because the
‘pictures’ of the numerals do not suggest the ‘thing’, but are themselves the
‘thing’. There was no need to go through the remaining three stages to make the
numerical alphabet more expressive or idealised. If we were to imagine an
alphabet or part of an alphabet built on these numeric ‘pictures’, only the last
stage would be necessary, i.e., assigning sounds to the ‘things’. If we were to
imagine an initial alphabet of ten shapes based on the ‘things’, adding more
sounds as the need arises is a simple matter of ‘modifying’ the ‘things’ by
means of reversing, flipping, rotating, shortening, lengthening, etc., to
produce endless letterforms.
It is commonly believed that early written symbols were based on
pictograms (also known as pictographs) and ideograms. These are the “picture
writing” and “ideographic or symbolic writing” Ego identified. A pictogram, as
everybody knows, is a symbol that portrays or represents an object or a concept
by illustration. Because the original ‘things’ are larger than can be reproduced
on stones and other surfaces, they are usually drawn smaller so that we have
miniature pictures of the original ‘things’. The study of scripts led some
experts to believe that pictograms appeared before ideograms, and they were used
around 5000 BC by ancient cultures in different parts of the world including the
Arabian Peninsula, Egypt and certain parts of the Mediterranean.
The inscription (below, Fig. 19) on the sarcophagus of Esmunazar, king of Sidon
(475-461) BC at the Musée du Louvre, Paris, contains four easily identifiable
shapes that resemble Arabic numeral,
8 (9, 3, 5, 8), which represent 40% of our
(above, Fiq, 20), appears to have three variations produced by reversing the numeral left
to right (
flipping it vertically and cropping the top part. Numeral (9)
appears to have two variations produced by curving the tail, and by adding an
extension to the circular line at the point it crosses the arched back. Thus,
the four original shapes were modified to provide five additional forms for a
total of nine characters. The original Canaanite alphabet is said to have
comprised 16 characters.
Character shapes vary from one inscription to another and they evolve
over time so some of the more recent shapes may appear very different from
earlier ones. Keeping this in mind, a survey of various inscriptions and texts
undertaken by Karl Faulmann (Das Buch der Schrift) from the texts available to
him yielded a surprising number of shapes that look similar not just to
eastern-style Arabic numerals but also to their western counterparts. If we add
character-mining work undertaken by other researchers, we can identify many of
the eastern and western-style Arabic numerals. Indeed, some Phoenician
ascriptions may look like an arithmetic exercise.
The early Aramaic alphabet reflects its strong Canaanite roots.
Extensive Aramaic inscriptions were found dating from the 10th century BC.
The 8th century BC silver ingot bearing the name of King
Bar-Rakib son of Panammu II, the king of Sam'al (modern Zenjirli in Turkey).
The orthography of the language in that early period was based on
Canaanite and the unity between the scripts of the two languages is clear on
comparison. Also clear in the inscription, read from right to left, the distinct
shapes of the Arabic numerals
(2), the reversed Arabic numeral
2 which is similar to the Arabic letter
(9) and its variants, and the reversed Arabic
It is up to the reader to decide if the presence of all these
numeral-shaped characters is circumstantial or whether many of the characters
are representations of the pictograms of Arabic numerals or their modified
shapes. Resolving the different varieties of letterforms to their primary forms
appears to lead us to some of the pictograms of finger and hand formations. If
we believe what we have been told by writers like Bowring, Gilchrist and others,
it is possible that most characters in the early Phoenician and Aramaic
alphabets could be derived from the ancient Arabian numeral pictograms.
Arab and Egyptian
Ancient Egyptian shares
with ancient Arabian “triconsonantal
roots, phemic-phonetic elements (such as the diversity of guttural sounds),
feminine nouns and second-person verbal forms ending in –t, gemination
(doubling) of the middle radical… similarities of the pronominal suffixes and
the independent personal pronouns, performative elements (such as the
S-causative and the N-reflexive), and many similarities in etymology.”
 To all this must be added one of the main
characteristics of languages related to the ancient mother tongue, i.e. the
right-to-left writing and reading of Hieratic and Demotic and the ability to
write and read hieroglyphic writing and numerals from right to left or
Many of the
clues to the origin of our numeral system are found in ancient Arabian, and the
original meaning of the number “one” in some of the descendent languages of that
great mother tongue is “finger”. How old the mother tongue of the ancient
Arabians is, again we cannot say. What is commonly known and
professed by experts, however, is that early written symbols were based on
pictograms that appeared around 5000 BC in different parts of the world
including the Arabian peninsula, Egypt and certain parts of the Mediterranean.
Our numerals are often described as “symbols”, “glyphs”, “forms”, “ideograms”,
and several other terms. However, they are not any of these. Except for the zero
which doubles as an ideogram, every numeral in our system is a pictogram, one of
the oldest forms of miniature drawings known to have been used by our ancestors
several thousand years ago. The system we often describe as “modern” may turn
out to be the oldest numeration system in history.
This may explain the familiar features of the Demotic alphabet as far as the
eastern shape-numerals of many of the letters on the stone are concerned. The
Demotic alphabet has more than 40 shapes with variant forms for
several letters, some of which are western-style numeral look-alikes.
Additional letterforms would have been needed, and these would have been made
available either by Arabians or by people in contact with them via Arabic
numeral pictograms that were used as phonograms.
Like the Canaanites who knew themselves by no other name, “Demotic” to the
Egyptians was an alien name of the popular language they identified as “sekh
shat” (writing for documents). Most of these documents were concerned with trade
and business in addition to the other fields for which languages are normally
used. Literature, science, religion, etc., are all important topics that seek
expression, but in the Middle East, the creation of literary and numeric scripts
seems to go hand in hand with the creation of wealth derived mainly from trade.
This was the case with the Phoenicians, South Arabians, Nabataeans, Palmyrenes, and Arabs. It is for these reasons that we thought it appropriate to
include in this book a chapter about trade. Indeed, the Arabic numeric system is
the creation of the marketplace in the same way that the system was taken to
Europe by traders, not by scholars and scientists.
The last two comments in this chapter are as follows: If the Arabic numerals are
ancient Arabian or Proto-Semitic as claimed, can we still call them “Arabic”?
The answer is not difficult. Of all other dialects, Arabic is the most loyal
daughter to a great mother tongue that we can only identify by matching its
features on the faces of its daughters. Compared to Canaanite, Aramaic, Arabic,
etc., very few Arabic numeral shapes are used in ancient Southern Arabian. This
may suggest that the Arabic numerals were used mostly in the northern part of
the Arabian Peninsula where they were picked up and reused by Nabataeans and
later by Arabs, whether via older Phoenician and Aramaic scripts, or by going
straight for the pictographic source available on the fingers and hands of
The numeral pictograms are a marvel on their own, but the Arabic numeric system
is primarily a cipher and positional system, and the concept of safr is
as Arabic as you can get. The fact that Arabs described locusts empty of eggs as
safr is not a reflection of their knowledge of biology, but of their want
for sustenance, because they used to supplement their meagre diet with locusts.
We presume they valued the eggs because without them, locusts were mere
exoskeletons unworthy of a meal. Locusts are still collected but sold mostly as
animal fodder, or so it is said.
To explain the source of the Arabic numeral forms in the Demotic script
of the Rosetta stone, we found it necessary to refer to distinguished writers
such as Breasted and Sir Alan to remind us of the extent of Semitic influence on
Egypt throughout its long history. The presence of the ancient Arabian numeral
pictograms and several manipulated variants in large quantity cannot be
attributed to chance. We are forced, therefore, to conclude that these numeral
shapes were essential constituents of the complex Demotic alphabet, upon which
much light needs to be shone before we can understand how it came into
existence, why and how the Arabic numeric language came to be incorporated into
the alphabet, and at what stage and from which source.
If the Canaanites used some of the ancient Arabian numeral pictograms
in their earliest alphabet or indeed at a later stage of its evolvement, we can
assume that these pictograms predate the Canaanite alphabet. If we add to this
the fact that numeral signs were invented before literary alphabet, then the
numeral shapes predate any known alphabet.
The tribal system in Arabia and elsewhere in the Middle East was not
simply a concentration of individuals but a matrix of culture and knowledge. As
such numeral shapes always existed in this matrix for innovative individuals to
pick and chose to create their first alphabets which consisted of 16 letters and
maybe less. It is, therefore, possible that with 10 numeral shapes available and
easily identifiable by a large number of people, they are very likely to be sufficient to create any
alphabet if the shapes are manipulated to provide easy identifiable letters.
The epitaph of king Imru'l Qais (Fig.27). Several letters have
the appearance of numeral shapes (Fig. 28) corresponding to the western numerals
Almost all classical Arabic references don’t differentiate between a
letter and a numeral. In recent times the word ‘raqam’ is used almost
exclusively to mean a ‘numeral’ or ‘number’. However, look up ‘raqam’ in
Lisan al-Arab and you’ll find that ‘a raqamised book’ is ‘a written book.’
Furthermore, raqam is also described as inscription and design ‘the origin of
which is writing.’ It seems, at an elementary level at least, that Arabs did not
find it important to differentiate between letter and number as if they had the
same origin. Moreover, several writers referred to numerals as ‘drawings’. We do
actually ‘draw’ our letters and numbers but as explained in Origin of the Arabic
Numerals, our numerals are pictograms (miniature drawings) of actual hand and
finger numeral formations.
The subject of ‘numeral-alphas’ is an open subject that will be
Edition of the Encyclopaedia Britannica – Numeral
Otto F. Ege, The Story of the Alphabet, (1921), p. 3.