Mnemonics for Arabic Vocabulary

Leveling up my Arabic is one of my projects for 2015. A particular point of difficulty with Arabic (at least for me, though not for native speakers) is the large number of consonant sounds, quite a few more than English. It follows that the closest English sound to a given Arabic consonant cannot be unique, including a surprising variety of sounds produced in the back of the throat.

Long ago, I was a guest in an Moroccan family's home. While I was there, something occurred which greatly upset the fifteen-year-old daughter of the house, who remonstrated loudly and vigorously with her parents. I sat there awestruck at the ability of her throat to handle this variety of scrapings and gluggings at high velocity with no apparent damage.

In some cases, the Arabic consonant is so alien to English that the mind gives up on finding an equivalent sound. Prominent example: the very word "Arab", in Arabic, begins with a consonant written ع, a kind of gagging sound. But apparently the first Europeans who borrowed the name balked at even trying to approximate it in their own language (whatever that was).

I will not be giving tips on Arabic pronunciation in this post, sorry. Please look elsewhere for such—I assume that you know at least the basics. This post will likely make no sense to you without at least a basic familiarity with Arabic script and phonetics.

I will, however, make a few remarks about Arabic grammar, because these are relevant to the working of the mnemonic system. One interesting aspect of Arabic (like other Semitic languages) is that words are generally composed of a root and a pattern, The root carries the core meaning, and is composed strictly of consonants, usually three. The pattern consists of vowels and sometimes additional consonants. The root and pattern are interleaved to make up the word.

We see a faint echo of this in English, if we consider, for example, the words "bear", "bore", and "burden." We could think of the core meaning as carried by the two consonants b-r. Combining this two-consonant root with patterns 1ea2, 1o2e, and 1u2den yields the three given words. Here I have described the patterns using numerals 1 and 2 to mark the position where the consonants would go.

Similarly, in Arabic, there is a root ك ت ب (k-t-b) which has to do generally with writing. Some of the words formed from this root by fitting it into different patterns would be:

كَتَبَ (kataba = k-t-b × 1a2a3a) "he wrote"
كِتَاب (kitaab = k-t-b × 1i2aa3) "book"
مَكْتَبَة (maktaba = k-t-b × ma12a3a) "library"

I hope my notation for combining root and pattern is fairly obvious. For example "k-t-b×1aa2i3" means insert the k, the t, and the b into the 1, 2, and 3 positions of the pattern, respectively, to yield kaatib "writer."

So recalling an Arabic word is a matter of recalling both the root (which usually consists of three consonants) and the pattern (which consists of vowels, possibly some additional consonants, with defined locations for inserting the consonants of the root). I use a different method for each of these.

First we deal with the root. The basic idea is to find an English word or phrase whose consonant sounds encode the three consonants of the Arabic root. Complications arise from the fact that Arabic has more distinct consonant sounds than English—this makes it impossible to simply use one English consonant for each Arabic consonant. Certain Arabic consonants such as ع or ء correspond closely to nothing in English, while sometimes more than one Arabic sound resembles a given English sound (such as س and ص, which both resemble English s). On the other hand, certain English consonants, such as p, have no close equivalent in Arabic.

The first step is to match each Arabic consonant with an English sound. I have made the two similar where possible—this is not essential, but it makes the system easier to learn. However, there are significant exceptions. By no means take these equivalents as a guide to Arabic pronunciation:

 ءN

ب b, B

ت t

ث th

ج J, SH, ZH

ح NG

خ ch, CH

د d

ذ TH

ر r

ز z

س s

ش j, sh, zh

ص S

ض D

ط T

ظ Z

ع g

غ G

ف f, v

ق K

ك k

ل L, l

م M, m

ن n

ه ng

و F, V

ي R

Note that I have assigned capital letters in some cases and small letters in others, in order to distinguish ق (K) from ك (k), although both Arabic letters sound at least something like English k. This raises the issue of keeping capital and small letters distinct, which we will deal with later. Note also: the th written above (whether capital or not) can be either the unvoiced th in "breath" or the voiced th in "breathe." I avoided using the English sounds w and y because these blur the lines between consonants and vowels: this is why Arabic ي, which sounds like y, was matched up with R instead. And I avoided using English h, because its pronunciation in certain words such as "wheel" comes and goes according to the speaker.

So the basic principle is to take the three letters of the Arabic root and generate the corresponding three English consonants. For example, the forementioned root ك ت ب (k-t-b in Arabic) turns into English k-t-b or k-t-B. (The k and t must be small but the b can be either capital or small.) (Note also that in ك ت ب the ك comes first and the ب comes last, because Arabic is written right-to-left.)

Or again the root ص ي ف, which has to do with "summer", would turn into S-R-f, or S-R-v, where the S and R must be capital and the f or v must be small.

The next step is to find an English phrase or word which contains the given sequence of consonant sounds. Here we must confront the issue of capital-versus-small letters, since these cannot be distinguished by sound. We handle this by appending a fourth English consonant sound which encodes the pattern of capital and small letters. There is a system to this: a combination of a binary coding and the Major system of  memorizing numbers, but with only eight patterns you don't need a system; it just makes it slightly easier to memorize if you understand both binary numbers and the Major system. There are eight possible patterns of capital (C) versus small (s), encoded thus:

sss: coded with s or z or f or v
ssC: coded with t or d or th or p or b
sCs: coded with n
sCC: coded with m
Css: coded with r
CsC: coded with l
CCs: coded with ch or sh or j or zh
CCC: coded with k or g

As before, these refer to sounds rather than letters.

So, now let's try some examples. As previously explained, the root ك ت ب, which relates to writing, translates to the English sequence k-t-b or k-t-B. Appending the corresponding pattern of capitals and smalls: sss (s, z, f, or v) or ssC (t, d, th, p, or b), we get a choice of English four-consonant sequences. We only need to represent any one of these with a word or phrase. I couldn't find a single word, but two-word phrases are easy: for example, k-t-B-t could be "cat bite" or "cut bait" or "cute boat" or "cat bathe", etc.

The root ص ي ف, which has to do with summer, translates to the English sequence S-R-f or S-R-v. Appending the corresponding pattern of capitals and smalls: CCs (ch, sh, j, or zh), we again get a choice of English consonant sequences, for example S-R-f-sh. And again I couldn't find a single English word containing all four consonants, but two-word sequences are easy: "sour fish", or "surf witch", etc.

I felt pretty good about this system, since it uses only four English consonants to encode three Arabic consonants. As mentioned previously, three English consonants do not suffice to encode three Arabic consonants, so in a sense this is the best attainable.

This leaves us with the question of how to represent the pattern, the complementary part of the Arabic word. Fortunately, this is a far simpler business, and I will describe it in an upcoming post.