LINGUIST List 6.1065

Thu Aug 10 1995

Sum: Antipassive and Reflexive

Editor for this issue: T. Daniel Seely <>


  1. Jeffrey L Lidz, Antipassive and Reflexive

Message 1: Antipassive and Reflexive

Date: Wed, 09 Aug 1995 21:07:58 Antipassive and Reflexive
From: Jeffrey L Lidz <>
Subject: Antipassive and Reflexive

Some time ago, I posted a question asking whether people knew of
languages that had both a reflexive morpheme (attached to V) and an
antipassive morpheme (attached to V), where the two were different. I
received three replies which are included below in abbreviated form.
I thank those who responded for their help: Angela Terrill, Mike
Darnell, and Suzanne Kemmer


There are thirteen or so Australian languages which have antipassives (all
Pama-Nyungan). Of these, most use the same morphology as their reflexives.
Some languages have antipassive verbal morphology but no reflexive
morphology. But the languages which have separate explicit verbal
morphology for reflexives and antipassives are, including sources:

* Dyirbal (around Cairns) marks antipassive with -ngay-. Reflexive is rriy
~ yirriy ~ marriy ~ (m)barriy.
Dixon (1972) The Dyirbal Language of North Queensland (Cambridge: CUP)

* Warungu (west of Dyirbal, not closely related) has antipassive -gali- and
reflexive -li-. Probably historically related.
T. Tsunoda (1974) 'A Grammar of the Warungu Language' Monash University MA

* Kalkatungu (western Queensland) has antipassive -yi- and reflexive -ti-.
Maybe historically related.
B. J. Blake (1979) A Kalkatungu Grammar (Canberra: Pacific Linguistics)

The proto-Pama-Nyungan (and maybe proto-Australian?) reflexive form is
something like *-DHi-rri-, where DH is a laminal stop, either
lamino-palatal or lamino-dental. The overwhelming majority of reflexive
(and antipassive) markers in Pama-Nyungan languages are reflexes of this.
The Warungu and Kalkatungu antipassive forms given above are probably
reflexes of this also, but the Dyirbal form definitely isn't (although,
interestingly, it is cognate with an applicative form in nearby Yidiny
(Dixon 1977).


Does this work for you?

lix-t-0-as ta sitn
put.dowm-trans-3sgO-3sgA art basket
He/she put down the basket.

lix-m-0 t-ta sitn
put.down-intrans-3sgS oblique case-art basket
He/she put down the basket.

Now admittedly there are some complexities to this situation. Some Salish
anists (the examples from Squamish, British Columbia) dispute the
antipassive analysis in the second example, but putting that aside

He/she prepared himself/herself.

This seems to fit your requirements, but I have a question.
Could the morpheme in your example, and perhaps in the case of others that
you mention in passing, be a middle voice marker rather than reflexive.
Again, from Squamish

suk'w-um (the -um here is identical to the -m above)
He/she bathes (only himself or herself, a transitive reading requires the
- -t, you saw above) Whatever the case with the examples that you've seen
Squamish seems to have a morpheme for antipassive and middle, with a
separate one for reflexive.


A number of Philippine languages have an antipassive
that consists in detransitivizing the verb with a prefix
and stripping the object of its 'determiner' (which marks
case) and other modifiers. The reflexive marker, on the
other hand, is a completely different morpheme.
Kapampangan is a language of this type.

Another case, I would suggest, is languages whose
former reflexive marker now codes antipassive, but
which in the meantime have innovated a new reflexive
marker. Russian uses the suffix '-sja' on the verb
'bite' to mean 'the dog bites' (generically) and
similar cases. But -sja is not the reflexive marker
in Russian. Sebja is the productive reflexive marker.
-Sja does occur on verbs like 'wash', but these are
not reflexive verbs -- languages tend to mark them
differently from reflexives. My book "The Middle Voice"
draws the reflexive-middle distinction and documents
it from many languages.

Jeff Lidz
University of Delaware Office: (302) 831-6489
Department of Linguistics Home: (302) 656-1902
46 E. Delaware Ave. Email:
Newark, DE 19716
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue