The Inuit language has a somewhat strange ("strange" to non-native speakers anyhow) grammatical feature called an "ergative" marker. First, I'll quote the Wikipedia entry on what ergativity means then try and explain as best as I can myself:
An ergative–absolutive language (or simply an ergative language) is a language in which the single argument ("subject") of an intransitive verb behaves like the object of a transitive verb, and differently from the agent ("subject") of a transitive verb. For instance, instead of saying "I moved her" and "she moved", speakers of an ergative language would say the equivalent of "I moved her" and "her moved". (http://en.wikipedia.org/wiki/Ergativity)
The question of "transitivity" is of primary importance in talking about "ergativity". The difference between 1) "I see" and 2) "I see him" is that 1 is "intransitive" (ie, grammatically complete without a direct object), and 2 is "transitive" (ie, specifies both the subject "I" and direct object "him"). The great subtlety is that the difference between the two phrases is not so much in the nouns but the verbs – though "to see" with respect to the two phrases in English seems, for all intents and purposes, unaltered by its grammatical function.
In the Inuit language, grammatical differences that have to do with "transitivity" are specified by its morpho-syntactic structure through the ergative markers. I say "markers" in the plural because most if not all the linguistics papers on the Inuit language that I've read seem to speak only of the pronominal endings (such as [-tara/-jara]; [-tait/-jait]; [-tanga/-janga]) but I noticed in thinking about Inuktitut that it seems possible to denote ergativity without having to mark it pronominally.
There is a particle in Inuktitut that seems to function like an ergative marker but nonetheless requires a direct object to complete its grammaticality (though strictly speaking, ergativity has to do with intransitive constructs). This particle is the [-si-] morpheme, as in:
kapisijuq "he stabs (someone)"
which is distinct from
kapijuq "he stabs (himself)"
This "complication" may have to do with the fact that some ergative languages have an additional grammatical function having to do with "active-stative" distinctions. Rather than trying to explain what "active-stative" means, let me just refer you to this Wikipedia entry (http://en.wikipedia.org/wiki/Active%E2%80%93stative_language).
I'm actually not joking. After going through all that complicated technical language in the entry above, the most important point has to do with this statement therein:
If the language has morphological case, then the arguments of a transitive verb are marked using the agentive case for the subject and the patientive case for the object, while the argument of an intransitive verb may be marked as either.
Clear as mud? Then, you're in good company (including yours truly, to be honest). But I think I can actually follow my own argument here, being a native speaker of Inuktitut with some technical serviceability in linguistics. But I also realize the trap I've willingly fallen into (namely, that abstract theoretical arguments have that frustrating aspect of seeming to want to count the number of angels dancing on the tip of a needle). At the risk of appearing snobbish I actually enjoyed falling into this trap, willingly.
Granted, the line of reasoning here is subtle but not all "aha!" moments translate well outside the subjective experience. I may not look it, but I'm actually as "happy as pig in...".