DOI Page / Citation link:
https://doi.org/10.11588/diglit.49270#0195
120
J. Abercromby
amiat, amiet, 'three’. Comparing this with amelotti, ameriat, 'three’ of List A it is evident that
an / or r has fallen out between the two vowels -ia-.
area, 'four’ is clearly a late loan word from (Ar.) arba’, ‘four’.
cansa, ‘five’, is the Ar. hamsa and again indicates that h was a foreign sound, or one that was
not liked. The interchange of m and n occasionally happens in Berber dialects, e. g. (Q.) anzar, (W.)
amzar, ‘rain’. (Q.) endi, (Bougie) emdi, ‘to set a snare’.
SUMUS, ‘six’, is the masc. form of simusetti, ‘five’, of List A and agrees with (Tam.) semmas, (Shil.)
summus, ‘five’.
sat, ‘seven’ corresponds with modern (Ar.) sitt, ‘six’, and with sesetti, ‘six’, of List A, (Tam.)
sediset, (Shil.) sddist, ‘six’.
set, ‘eight’ is the satti, ‘seven’ of List A; (Shil.) ssat, ‘seven’.
acot, ‘nine’, is the masc. form of acodctti, ‘four’, of List A; (Tam.) okkoz, (Shil.) (kuz), ‘four’.
marago, ‘ten’, is the marava of List A. In Berber dialects g sometimes results from the contraction
of a double u, e. g. (Tam.) iggat, ‘he strikes often’ is for iuuat, from iuot, ‘he struck’. (Mz.) aggai,
‘a burden’ from aui, ‘to carry’. But it can also develop from a single u, e. g. (Zen.) tutfifl, (W.) tagdefit,
‘an ant’. In this way marava has become marago.
beni marago, ‘eleven’. In the Berber dialects a copula d, ‘and ’, is inserted between the units and
the tens, as will be noticed in List C. Tn the Canarian dialect it appears to be omitted, as is also the case
in Hamitic Bedja.
lini marago, ‘twelve’, calls for no remark.
linago, ‘twenty’, amiago, ‘thirty’, arbiago, forty, cansago, fifty, etc., as far as acotago, ninety
are very remarkable. The tens are formed by bisecting the word for ten, mar-ago, and using the second
half, which remains in the singular, as a suffix. Such a procedure is cpiite contrary to Berber usage and
there is no analogy for it.
beemaragoin, ‘hundred’, maragoin is the plural of marago and is quite conformable with (Mz.)
merau-in, ‘tens’, (Tam.) temeru-in, (Shil.) temer-aw-in, ‘a set of ten’, [(Fr.) dizaine], with which the tens
are formed in the Sahara and Morocco. But it is evident ‘one (set of) tens’, bee maragoin, does not
itself mean ‘hundred’. As the Beni Mzab sometimes say merau merauin, ‘ten tens’ for a hundred, it
may be supposed that bee maragoin is shortened from bee marago maragoin ‘one ten-tens’, the middle
word having been omitted in course of time for the sake of brevity.
These two lists, A and B, when compared show that both have native words only as far as four,
though in List B the old word acot, ‘four’ has been displaced and removed to nine for a reason which
will be explained below. From five to eight inclusive the words have been borrowed from the Arabic,
apparently not directly but through the medium of the adjacent Berbers of the mainland. The words
for seven and eight of List A make this nearly certain.
Though this borrowing, which is discernible in List A, is relatively old, it does not take us very far
back, and we are left in the dark as regards the method of counting in the really early times when the
Canary Islands received their first human population from the opposite coast. It is, however, possible
that the natives used a quinary and not a decimal system. If so they would not stand alone, for traces
of a quinary system have been observed in the Hamitic languages of the Bedja, Bilin, and Khamir.
Basset mentions that in the dialects of Mzab, Gebel Nefusa, and Wad Rig the natives sometimes say
afus, ‘hand’ for five, afus cl iggen ‘hand and one’ for six, and so on.
A trace of this quinary system may be perceived in List B, for it helps to account for some of the
displacements in the value of the names. As Dr. Berthelot has remarked, the words for four and five
in List A correspond with nine and six of List B. Both he and Pietschmann supposed that this was
owing to a mistake on the part of the compiler of the list. Another explanation, however, presents
itself as a possibility which is at least worthy of consideration.
The aldamorava, ‘nine’, of List A has no relation to (Tam.) tessaa, tessahat, which is evidently bor-
rowed from (Ar.) tis‘-u, fem. tis'-at-u, ‘nine’, and has no place in List B, When the new words arba,
J. Abercromby
amiat, amiet, 'three’. Comparing this with amelotti, ameriat, 'three’ of List A it is evident that
an / or r has fallen out between the two vowels -ia-.
area, 'four’ is clearly a late loan word from (Ar.) arba’, ‘four’.
cansa, ‘five’, is the Ar. hamsa and again indicates that h was a foreign sound, or one that was
not liked. The interchange of m and n occasionally happens in Berber dialects, e. g. (Q.) anzar, (W.)
amzar, ‘rain’. (Q.) endi, (Bougie) emdi, ‘to set a snare’.
SUMUS, ‘six’, is the masc. form of simusetti, ‘five’, of List A and agrees with (Tam.) semmas, (Shil.)
summus, ‘five’.
sat, ‘seven’ corresponds with modern (Ar.) sitt, ‘six’, and with sesetti, ‘six’, of List A, (Tam.)
sediset, (Shil.) sddist, ‘six’.
set, ‘eight’ is the satti, ‘seven’ of List A; (Shil.) ssat, ‘seven’.
acot, ‘nine’, is the masc. form of acodctti, ‘four’, of List A; (Tam.) okkoz, (Shil.) (kuz), ‘four’.
marago, ‘ten’, is the marava of List A. In Berber dialects g sometimes results from the contraction
of a double u, e. g. (Tam.) iggat, ‘he strikes often’ is for iuuat, from iuot, ‘he struck’. (Mz.) aggai,
‘a burden’ from aui, ‘to carry’. But it can also develop from a single u, e. g. (Zen.) tutfifl, (W.) tagdefit,
‘an ant’. In this way marava has become marago.
beni marago, ‘eleven’. In the Berber dialects a copula d, ‘and ’, is inserted between the units and
the tens, as will be noticed in List C. Tn the Canarian dialect it appears to be omitted, as is also the case
in Hamitic Bedja.
lini marago, ‘twelve’, calls for no remark.
linago, ‘twenty’, amiago, ‘thirty’, arbiago, forty, cansago, fifty, etc., as far as acotago, ninety
are very remarkable. The tens are formed by bisecting the word for ten, mar-ago, and using the second
half, which remains in the singular, as a suffix. Such a procedure is cpiite contrary to Berber usage and
there is no analogy for it.
beemaragoin, ‘hundred’, maragoin is the plural of marago and is quite conformable with (Mz.)
merau-in, ‘tens’, (Tam.) temeru-in, (Shil.) temer-aw-in, ‘a set of ten’, [(Fr.) dizaine], with which the tens
are formed in the Sahara and Morocco. But it is evident ‘one (set of) tens’, bee maragoin, does not
itself mean ‘hundred’. As the Beni Mzab sometimes say merau merauin, ‘ten tens’ for a hundred, it
may be supposed that bee maragoin is shortened from bee marago maragoin ‘one ten-tens’, the middle
word having been omitted in course of time for the sake of brevity.
These two lists, A and B, when compared show that both have native words only as far as four,
though in List B the old word acot, ‘four’ has been displaced and removed to nine for a reason which
will be explained below. From five to eight inclusive the words have been borrowed from the Arabic,
apparently not directly but through the medium of the adjacent Berbers of the mainland. The words
for seven and eight of List A make this nearly certain.
Though this borrowing, which is discernible in List A, is relatively old, it does not take us very far
back, and we are left in the dark as regards the method of counting in the really early times when the
Canary Islands received their first human population from the opposite coast. It is, however, possible
that the natives used a quinary and not a decimal system. If so they would not stand alone, for traces
of a quinary system have been observed in the Hamitic languages of the Bedja, Bilin, and Khamir.
Basset mentions that in the dialects of Mzab, Gebel Nefusa, and Wad Rig the natives sometimes say
afus, ‘hand’ for five, afus cl iggen ‘hand and one’ for six, and so on.
A trace of this quinary system may be perceived in List B, for it helps to account for some of the
displacements in the value of the names. As Dr. Berthelot has remarked, the words for four and five
in List A correspond with nine and six of List B. Both he and Pietschmann supposed that this was
owing to a mistake on the part of the compiler of the list. Another explanation, however, presents
itself as a possibility which is at least worthy of consideration.
The aldamorava, ‘nine’, of List A has no relation to (Tam.) tessaa, tessahat, which is evidently bor-
rowed from (Ar.) tis‘-u, fem. tis'-at-u, ‘nine’, and has no place in List B, When the new words arba,