DOI Page / Citation link:
https://doi.org/10.11588/diglit.49270#0194
Canarian Speech
119
List A
nait, ‘one’. In Zenaga there are two forms, iun and niu. The feminine is not given by Faid-
herbe, but *niut might be expected, which compares well with nait.
smetti, ‘two’ with the change m into n, which occurs in some Berber dialects, smetti compares well
with (Shil.) snat, (M.) senat, snat.
The doubling of the t in all the words from two to eight inclusive is not easy to account for. It
might be the copula de, di, ‘ and ’, which would become ti when following t, supposing that the native
said ‘two and’, ‘three and’ etc., when counting.
amelotti, ‘three’. The I is uncertain as the word for thirteen is ameriat marava. But whether
I or r is the correct consonant the word does not correspond with the expression for ‘ three ’ in any
Berber dialect.
acodetti, ‘four’, is the equivalent of (Tam.) okkozet, (M.) akkozt. Here the d may be earlier
than the z. Examples of this change are (Tam.) ebdeg, (Zen.) udag, (Q.) ebzeg, ‘to be wet’; (Tam.)
egged, (Q.) neggez ‘to jump’.
simusetti, ‘five’, is (Tam.) semmuset, (M.) semmust. The word has evidently a close affinity
with (Ar.) hamsat-un, five. It is in all probability an early loan word, but as h is not an original sound
in Berber it was replaced by s. The initial s indicates that the Canarian word was taken over from the
Berbers of the mainland and not directly from Arabs.
sesetti, ‘six’, agrees closely with (Tam.) seddiset, (M.) saddist, though the d has dropped out.
In the (Ar.) sitt-at-un, ‘six’, the tt is a contraction and assimilation of dis, which is preserved in the ordinal
sadis-at-un, ‘sixth’. This indicates that the Berber loan word was adopted at an early date, not improb-
ably before the era of Muhammad when Arabic first began to be written in books.
satti, ‘seven’, (Tam.) essahat, (M.) sat. All these are contracted forms of (Ar.) sabatun, ‘seven’.
tamatti, ‘eight’, (Tam.) ettamet, (M.) tamt. All these are shortened from (Ar.) Oamaniyat-un,
‘eight’. The great attrition that this word and the word for seven have undergone is proof that they
have been very long in use and were borrowed as early as sesettib, ‘six’.
aldamorana, ‘nine’'. As Berthelot remarked, the n is doubtless a misread v and m'orana is the
same as marava. The Canarian word has no connection with the Berber word for nine, and the alda is
difficult to explain. Cf. (Tait.) illad i, ‘ on this side of’. It governs the preposition i, ‘to’, e. g. illad i
teflut, ‘on this side of the door’. So iZZud i morava would mean ‘on this side of ten’. Or cf. shawia alda,
‘as far as, up to’; alda morava, ‘as far as, up to ten’.
marSva, ‘ten’ corresponds with (Tam.) merau, but it is an older form as it retains a final a. It
will be observed that marava is masculine and not feminine like the words from one to eight inclusive.
The numerals in the list from eleven to sixteen require no particular comment, though it will be
observed that the units precede the tens as in the Maroccan dialects, in Zenaga, and in Arabic, though
not in Tamoseq. The difference in the vowels and the absence of final ti in three of the words may be
explained by the negligence or inadvertence of the scribe so that nothing definite can be built upon them.
List B
been, ben, ‘one’. In Spanish books of the 16th and 17th centuries v is often used for w, so ben may
be another way of writing uen, a form which compares with (W., Zen.) iun, (Q.) iuen, ‘one,’ though we
have in (Tam.) iyen, in (Mz.) iggen.
lini, ‘two’. Although in the Hamitic languages somewhat similar forms are to be found, as in
(Bilin) lana, (Khamir) liha, ‘two’, (Galla) lama, (Bedja) malo, ‘two’, it is preferable to suppose that
lini is a dialectical form of a Berber word. The (Tam.) sin, ‘two’, is possibly for an older *snin, cf.
(Ar.) iflnani, ‘two’, in which case an old Canarian dialect may have dropped initial 5 and retained- nin.
The interchange of n and I is possible phonetically, both being liquids, and occurs, though rarely, in
Berber, e. g. (Mz.) tisubna, (W.) tsubla ‘a large needle’, (Tait.) tissubla ‘an awl’.
119
List A
nait, ‘one’. In Zenaga there are two forms, iun and niu. The feminine is not given by Faid-
herbe, but *niut might be expected, which compares well with nait.
smetti, ‘two’ with the change m into n, which occurs in some Berber dialects, smetti compares well
with (Shil.) snat, (M.) senat, snat.
The doubling of the t in all the words from two to eight inclusive is not easy to account for. It
might be the copula de, di, ‘ and ’, which would become ti when following t, supposing that the native
said ‘two and’, ‘three and’ etc., when counting.
amelotti, ‘three’. The I is uncertain as the word for thirteen is ameriat marava. But whether
I or r is the correct consonant the word does not correspond with the expression for ‘ three ’ in any
Berber dialect.
acodetti, ‘four’, is the equivalent of (Tam.) okkozet, (M.) akkozt. Here the d may be earlier
than the z. Examples of this change are (Tam.) ebdeg, (Zen.) udag, (Q.) ebzeg, ‘to be wet’; (Tam.)
egged, (Q.) neggez ‘to jump’.
simusetti, ‘five’, is (Tam.) semmuset, (M.) semmust. The word has evidently a close affinity
with (Ar.) hamsat-un, five. It is in all probability an early loan word, but as h is not an original sound
in Berber it was replaced by s. The initial s indicates that the Canarian word was taken over from the
Berbers of the mainland and not directly from Arabs.
sesetti, ‘six’, agrees closely with (Tam.) seddiset, (M.) saddist, though the d has dropped out.
In the (Ar.) sitt-at-un, ‘six’, the tt is a contraction and assimilation of dis, which is preserved in the ordinal
sadis-at-un, ‘sixth’. This indicates that the Berber loan word was adopted at an early date, not improb-
ably before the era of Muhammad when Arabic first began to be written in books.
satti, ‘seven’, (Tam.) essahat, (M.) sat. All these are contracted forms of (Ar.) sabatun, ‘seven’.
tamatti, ‘eight’, (Tam.) ettamet, (M.) tamt. All these are shortened from (Ar.) Oamaniyat-un,
‘eight’. The great attrition that this word and the word for seven have undergone is proof that they
have been very long in use and were borrowed as early as sesettib, ‘six’.
aldamorana, ‘nine’'. As Berthelot remarked, the n is doubtless a misread v and m'orana is the
same as marava. The Canarian word has no connection with the Berber word for nine, and the alda is
difficult to explain. Cf. (Tait.) illad i, ‘ on this side of’. It governs the preposition i, ‘to’, e. g. illad i
teflut, ‘on this side of the door’. So iZZud i morava would mean ‘on this side of ten’. Or cf. shawia alda,
‘as far as, up to’; alda morava, ‘as far as, up to ten’.
marSva, ‘ten’ corresponds with (Tam.) merau, but it is an older form as it retains a final a. It
will be observed that marava is masculine and not feminine like the words from one to eight inclusive.
The numerals in the list from eleven to sixteen require no particular comment, though it will be
observed that the units precede the tens as in the Maroccan dialects, in Zenaga, and in Arabic, though
not in Tamoseq. The difference in the vowels and the absence of final ti in three of the words may be
explained by the negligence or inadvertence of the scribe so that nothing definite can be built upon them.
List B
been, ben, ‘one’. In Spanish books of the 16th and 17th centuries v is often used for w, so ben may
be another way of writing uen, a form which compares with (W., Zen.) iun, (Q.) iuen, ‘one,’ though we
have in (Tam.) iyen, in (Mz.) iggen.
lini, ‘two’. Although in the Hamitic languages somewhat similar forms are to be found, as in
(Bilin) lana, (Khamir) liha, ‘two’, (Galla) lama, (Bedja) malo, ‘two’, it is preferable to suppose that
lini is a dialectical form of a Berber word. The (Tam.) sin, ‘two’, is possibly for an older *snin, cf.
(Ar.) iflnani, ‘two’, in which case an old Canarian dialect may have dropped initial 5 and retained- nin.
The interchange of n and I is possible phonetically, both being liquids, and occurs, though rarely, in
Berber, e. g. (Mz.) tisubna, (W.) tsubla ‘a large needle’, (Tait.) tissubla ‘an awl’.