Arabic ’alladī as a demonstrative

While I was reading through Li Guo’s edition of the mercantile letters from Quseir, Egypt, I noticed something interesting about the use of the Arabic terms ’alladī. In Classical Arabic, ʾalladī is an inflected pronoun that is used to introduce a relative phrase after a determinate antecedent, so for example you would say: al-rajul ʾalladī raʾaytu-hū fī l-sūq, ‘the man whom I saw in the market’. In some few cases, this relative pronoun can be elided, such as in the Qurʾān: ka-matal l-ḥimār yaḥmilu l-ʾasfār, ‘[they are] like donkeys [who] carry books’.

It’s been pretty well-described that in some cases, the relative pronoun ʾalladī was reanalyzed as a conjunction instead, such as in the following Classical Arabic phrase:

al-ḥamdu          li-llāh               ʾalladī ḫalaqa                l-nās
def-praise       prep-god         rel       he.created
‘praise be to God, who created mankind’

Which can be reanalyzed as:

al-ḥamdu          li-llāh               ʾalladī   ḫalaqa           l-nāS
DEF-praise       PREP-God         CONJ.  he.created
‘praise be God, that he created mankind’

Examples of this happening are all over the place: we can find it in Judaeo-Arabic letters from the 13th century (for example, in the phrase fəraḥnā ʾalladī ğamaʿ llāh šəlāmnā fī məkān wāḥəd, ‘we are glad that God put us all safely(?) in one place), as well as currently in the spoken dialects of Egypt and the Levant (as Phillip Stokes pointed out: ’anā mabsūṭ ʾillī ğīt, ‘I’m glad that you came’), so I’m not going to really spend more time discussing this.

So we know that Arabic ʾalladī can be used to 1) introduce the relative phrase (as in Classical Arabic) and 2) as a conjunction (in Middle Arabic and the Arabic dialects). However, these two functions do not seem to be sufficient to explain its usage in these two 13th century documents from Qusayr. Let’s look at what they say:

  1. wa-mā sayyarnā hādā ʾillā tasāʿdnā ʿalā ʾalladī fa-llāh allāh (135-36)
    ‘and we will not sent it unless you help us with [ʾalladī]. Oh God, oh God […]’
  1. ʾalladī ʾaʿlam bih al-ʾaḫ ʾaḥmad ʾan […] (240-41)
    ‘[ʾalladī] I inform the brother of Ahmad this, that […]’


What’s odd about these two phrases is that ʾalladī can not really be taken here as a relative pronoun: in example 1 because there is no following determining clause. The editor argues that it is a case of ellipsis, in that ʾalladī actually refers back to what is being discussed in the letter (a shipment of some sort), but I’m not so convinced. I think in this case it would be more logical to expect a resumptive pronoun with ʿalā. Considering the context, it’s not likely that ʾalladī was being used as a conjunction either, considering how this is the end of the sentence (as indicated by the particle fa-).

As for example 2, if ʾalladī was supposed to introduce a relative clause, then where is the antecedent? The letter itself starts with ʾalladī, so I’m not really sure what it’s supposed to refer to. A conjunction is similarly unlikely, considering the presence of the pronominal suffix -h after ʾaʿlam.

So my suggestion is that the use of ʾalladī reflects another function, that of a demonstrative pronoun (maybe distal, maybe proximal?), so we could translate the two phrase as following:

  1. wa-mā sayyarnā hadā ʾillā tasāʿdnā ʿalā ʾalladī fa-llāh allāh
    ‘and we will not sent it unless you help us with this/that Oh God, oh God.
  2. ʾalladī ʾaʿlam bi-h al-ʾaḫ ʾaḥmad ʾan […]
    ‘This/that, that I inform the brother Ahmad of that he’ […]

There are, of course, problems with this hypothesis. Firstly, I’ve only encountered ʾalladī like in this particular environment, and mercantile letters can reflect a specific kind of linguistic usage that is not representative of how people actually speak. Secondly, the corpus is rather thin, so I would definitely need more examples to make a stronger case about this. However, of two things I’m convinced: firstly, you cannot have a relative pronoun or a conjunction at the very end of a phrase; nor at the very beginning.