Consonants of Standard Arabic

The consonantal inventory of Arabic is quite daunting for a speaker of an Indo-European language.  There is a large number of guttural (velar/uvular/pharyngeal/glottal sounds).  Modern Hebrew does not pose such a problem for two reasons.  Firstly, Hebrew had already in antiquity lost some of the consonants of Proto-Semitic, notable some of the ephatic consonants.  Secondly, Modern Hebrew has since lost almost all of the remaining difficult consonants, including the pharyngeals and all the emphatics.  This is likely due to influence from Indo-European languages such as Yiddish.

I always get confused when I try to remember what all the sounds of Arabic are.  Here is attempt to classify them.  I shall ignore the relatively easy sounds of /m, b, f, w, θ, ð, n, t, d, s, z, l, r, d͡ʒ and ʃ/ which all appear in Standard English (except for /r/, which appears however in many other IE languages).   I shall concentrate on the more difficult guttural consonants.

Using the IPA as a basis, we get the following table (the phones in bold occur in MSA)

	velar	uvular	pharyngeal	glottal
nasal	ŋ	N
stop	k g	q G		ʔ
fricative	x ɣ	χ ʁ	ħ ʕ	h ɦ
approximant	ɰ	(ʁ)	(ʕ)

/x/ and /ɣ/ aren’t too tricky to pronounce: /x/ appears in the German Bach and /ɣ/ in the Spanish agua.  You will notice that according to IPA, the distinction between fricative and approximant is lost in the uvular and pharyngeal columns.  This makes sense to me.  Even in the velar column, /ɣ/ and /ɰ/ sound very similar to me.  /q/ is also not that hard to pronounce, merely being a /k/ pronounced as far back in the throat as you can.  /h/ is an English phoneme, as is /ʔ/, albeit marginally, in uh-oh and the Cockney bottle.

The really difficult ones are /ħ and ʕ/ which honestly just sound like voiced and unvoiced choking sounds.  Technically, they are produced by pharyngeal constriction.  In Arabic, the uvular and velar columns are more alike (/x and χ/ are allophones, as are /ɣ and ʁ/).  However it makes more sense for me to think of the velar consonants as being, well, velar, and to group the uvular and pharyngeal consonants together, concentrating on producing them as far back in the throat as possible.  Ignoring the straightforward /h/ and /ʔ/, this leaves:

	velar	uvular/pharyngeal
stop	k g	q G
fricative	x ɣ	ħ ʕ

Which is not particularly hard to remember.  And now describing the emphatic consonants /tˤ dˤ sˤ  and zˤ~ðˤ/ becomes reasonably straightforward too, with /tˤ/ being roughly equivalent to coarticulated /t + ħ/, /dˤ/ to /d + ʕ/, and so on.

And here is some vocab for comparison.

Numbers

	MSA	Moroccan A	Egyptian A	Mod Hebrew
1	waaħid	waaħed	wāħed	ehad
2	iθnaan	ʒuʒ	ʔetnēn	ʃnayim
3	θalaaθa	tlata	talāta	ʃloʃa
4	arbaʕa	rebʕa	ʔarbaʕa	arbaʔa
5	xamsa	xemsa	xamsa	hamiʃa
6	sitta	setta	setta	ʃiʃa
7	sabʕa	sebʕa	sabʕa	ʃivʔa
8	θamaaneya	tmenya	tamanya	ʃmona
9	tisʕa	tesʕuud	tesʕa	tiʃʔa
10	ʕaʃara	ʕaʃara	ʕaʃra	ʔassara

Arabic consonants – derivation + writing

You may have noticed that there are quite a few extra letters in the Arabic language, compared to the basic Semitic alphabet.  These are derivatives of the basic letters:

• ذ ð
• خ x
• ظ ẓ/ðˤ
• غ ɣ
• ض ḍ
• ش ʃ
• ث θ

There are quite a few modified Hebrew letters too, but these exist mostly to allow the writing of non-indigenous phones, such as /p/ and /v/.  The Semitic alphabet actually fits the Hebrew language much better than the Arabic language, because it was designed to transcribe North-West Semitic languages (i.e. Canaanite, Phoenician) whose modern descendent is Hebrew.  Arabic is a Central Semitic language which has retained the above consonants from Proto-Semitic.

Proto-Semitic had 5 triads of voiced/voiceless/emphatic consonants:

• Dental stops: d t tˤ
• Velar stops: g k kˤ (phonetically g k q)
• Interdental: ð θ θˤ
• Dental sibilants: z s/ʃ sˤ
• Lateral: l ɬ ɬˤ (also written l ś śˤ)

As Proto-Semitic evolved into North-Western Semitic, several of these consonant sounds coalesced.  The dental stops remained separate (although in modern Hebrew, taw t and teth tˤ are pronounced the same), as did the velar stops.  Among the sibilants (dental, interdental and lateral):

• the voiced ð and z coalesced to make zayin z, but l lamed remained separate.
• the unvoiced θ, ʃ and ɬ coalesced to make shin ʃ, but s samekh remained separate.
• the emphatic θˤ, sˤ and ɬˤ coalesced to make sadhe sˤ (modern hebrew pronounced ts).

The case of shin is the only case where Ancient Hebrew and earlier North-Western languages such as Phoenecian where not in alignment.  While the Ancient Hebrew script followed the Phoenician example, having only one letter for the coalesced θ, ʃ and ɬ, in the spoken language, θ and ʃ had coalesced to ʃ while ɬ had transformed into s.  Later diacritics were invented to mark this difference.

The Central Semitic languages such as Arabic retained all the distinctions of the Proto-Semitic language, except for the loss of samekh.  By and large the Arabic sounds are the same as those in Proto-Semitic.  Exceptions include:

• ɬ -> ʃ and ʃ -> s (the opposite of Hebrew)
• θˤ -> ðˤ~zˤ
• ɬˤ -> dˤ

The origin of the “extra” fricative letters in Arabic is obscured by the fact that these letters are largely not based on the letters with which the sounds coalesced in North-West Semitic, but on the letters which they most ressemble phonetically, hence:

• ð is spelt ذ, derived from daleth, not zayin (ð -> zayin in Hebrew)
• θ is spelt ث, derived from taw, not shin (θ -> shin in Hebrew)
• ðˤ~zˤ (PS θˤ) is spelt ظ , derived from teth, not sadhe (θˤ -> sadhe in Hebrew)

But the following letters do reflect derivation:

• ʃ (PS ɬ) is spelt ش, derived from shin (Arabic ʃ corresponds to shin in Hebrew)
• dˤ (PS ɬˤ) is spelt ض. , derived from sadhe (Arabic dˤ corresponds to sadhe in Hebrew)

Proto-Semitic also had a voiced and unvoiced pair of velar/uvular and pharyngeal fricatives:

• voiced: ɣ/ʁ, ʕ
• unvoiced: x/χ, ħ

These coalseced on the basis of voicing in North-Western Semitic, but not in Central Semitic.

PS	North-West Semitic		Central Semitic
	Phoen.	MH	MSA
t	t taw	t	t ت
tˤ	tˤ teth	t	tˤ ط
d	d daleth	d	d د
ð z	z zayin	z	ð ذ from daleth z ز
l	l lamed	l	l ل
θ ʃ ɬ	ʃ shin	ʃ ʃ s	θ ث  from taw s س ʃ ش
s	s samekh	s	–
θˤ ʃˤ ɬˤ	sˤ sadhe	ts	ðˤ~zˤ ظ  from teth sˤ ص dˤ ض
ʕ ɣ/ʁ	ʕ ayin	ʔ	ʕ ع ɣ غ
ħ x/χ	ħ heth	χ	ħ ح x خ

The Semitic alphabets

When I tried to learn the Arabic alphabet a few years ago I found it really hard because the ordering of the letters seems very haphazard.  However there is an alternate method of ordering which follows the original Semitic (e.g. Phonecian) alphabet.  It resembles the Greek and Latin alpabets closely.

Ph	Hebrew		pron (if dif)	Arabic		pron (if dif)	vars	Greek	Latin
/ʔ/	aleph	א	/Ø,ʔ/	‘alif	أ	/Ø,ʔ/		alpha	A
/b/	beth	ב		bā’	ب			beta	B
/g/	gimel	ג		jīm	ج	/dʒ/		gamma	C G
/d/	daleth	ד		dāl	د		ذ ð	delta	D
/h/	he	ה		hā’	ه			epsilon	E
/w/	waw	ו	/v (w)/	wāw	و			digamma, upsilon	F U V W Y
/z/	zayin	ז		zayn	ز			zeta	Z
/ħ/	heth	ח	/χ(ħ)/	ḥā’	ح		خ x	eta	H
/ṭ/	teth	ט	/t/	ṭā’	ط		ظ ẓ/ðˤ	theta
/j/	yod	י		yā’	ي			iota	I
/k/	kaph	כ/ך	/χ/	kāf	ك			kappa	K
/l/	lamed	ל		lām	ل			lambda	L
/m/	mem	מ/ם		mīm	م			mu	M
/n/	nun	נ/ן		nūn	ن			nu	N
/s/	samekh	ס						xi, chi	X
/ʕ/	ayin	ע	/ʔ(ʕ)/	‘ayn	ع		غ ɣ	omicron	O
/p/	pe	פ/ף	/f/	fā’	ف	/f/		pi	P
/ṣ/	sadhe	צ/ץ	/ts/	ṣād	ص		ض ḍ	san
/q/	qoph	ק	/k/	qāf	ق			koppa	Q
/r/	resh	ר	/ʁ/	rā’	ر			rho	R
/ʃ/	shin	ש		sīn	س	/s/	ش ʃ	sigma	S
/t/	tav	ת		tā’	ت		ث θ	tau	T