The pillars of eternity world is known of its many conlangs and the latest entry in the setting, Avowed, is no exception. And while the wiki seems pretty extensive about most of the languages featured in the games, the language of the Ekida (Ekidan from now on), the "godless people", seems to be an exception. Maybe it is because Avowed was a critical failure, maybe it is because the game does not make a great point of presenting it. Either way, someone needs to do something about it, and I propose that that someone be me.
This page will be dedicated to my research of this language as I play through the game. I will also be compiling a corpus of all the Ekidan I come across.
Once I consider my work done, I'll share my findings on the wiki, but as it is, I feel like keeping my drafts to myself. I'll try to keep my work here as tidy as possible, but it might degenerate into a blog style/chronological exposé if I become lazy.

Ekidan Corpus:

Entry	Meaning	Metadata	Notes	ID
Yul hesiza Yezuha lu yedez rageda me gali hed sayapayi Enwitan. K'gid yul gapa pubi.	Not given an in-game translation.	Found in the temple of Eothas. Engraved on clay tablet in an unknown writing system. Still readable by the player.	Might not be Ekida, further testing might be needed.	LR01
yiz	Possibly "god"	Found in History of the living lands, vol. IV, censored	Unreliable source, might not be Ekidan.
Naku Tedek	"Ancient Soil"	Toponym	Translation given by Giatta
Ekida	Ethnonym
Sapadal	God of the living lands and the Ekida	Theonym	Came to the player in a spiritual communication with a god, under the influence of concentrated essence, during an animancy experiment led by Giatta.
Naku Kubel		Toponym
Enhekala		Idionym	Legendary figure
Nekkad		Idionym	Legenday figure
Tebara Sanakis		Theonym	No literal translation given (yet), probably not purely idiosyncratic.

Analysis of recurring pattern in toponyms:
So far, I've encountered to Ekidan toponyms: Naku Tedek and Nalu Kubel. Only one of them as been translated: Naku Tedek, "Ancient Soil".
It is relatively safe to assume, for now, that there is a 1:1 correspondance for the english and Ekidan name. While it is yet impossible to determine the exact meaning of Naku Kubel, the pressence in both toponyms of Naku is a nice start.
Once I know what Naku Kubel means, we will have a minimal pair, granting us the meaning of three words.

On "Yul hesiza Yezuha lu yedez rageda me gali hed sayapayi Enwitan. K'gid yul gapa pubi.".
While I do not know for sure if this is Ekidan, and don't have any translation, it is still possible to do some preparatory analysis.
First, let's look at the capitalized words, probably proper nouns:
Yezuha: according to the Pillars of Eternity wiki, Yezuha is a "mysterious region or nation to the east of the Deadfire, beyond the storms of Ondra's Mortar.". They speak a language called Seki. I don't know if this inscription is Seki or if it was written contemporary to the existance of the Yezuha nation.
Enwitan: seems to be this language's version of "Engwithan". This informs us as to the subject of the text: it is probably an historical account or a quote regarding the Yezuha et Engwithan. Linguistically speaking, this informs us that the language of the inscription doesn't have the sound corresponding to "th" in Engwithan or that the this sound is unstable in this position. Since I have yet to find this sound in any Ekida writing, I surmise it just doesnt exist and has been related to /t/ in the loaning process.
The loss of the "g" in Enwitan is more problematic since "g" seems common in Ekidan, even in this very inscription. However, this inscription does not contain any consonnantic cluster inside words (except for K'g, which we think represents of velarized /g/? or a specific realization of /g/ or /k/). This leads me to believe that, while /g/ is possible in Ekidan, its position in Engwithan: C_G, which is not attested in Ekidan, led to its disparition. EDIT: I have found a CC consonant cluster in "Enhekala", it's still not a CCG or CCC consonant cluster but it does provide us with more phonology context.

Only one word is repeated in this inscription: "yul". Both times it is placed at the beginning of the sentence and it is the only word present in both sentences. These simple facts alone cannot inform us as to the meaning of "yul", but is tells us to stay on the look out for it in other texts. It could indicated that is be a grammatical word rather than a lexical word, but I should be careful with this kind of assumption.

I've finally reached the point of the game where Yatzi is in my party and i can talk to her about her research. Honestly, it's been quite lackluster on the linguistic front so far but she did teach me something about the history of the Ekidan that could give us some information, depending on the work the writers put into making their languages coherent.
So, what exactly did I learn?
It seems, as I had gleaned from inscription LR01, the ekidan are linked both to Yezuha and the Engwithan, more precisely, it seems the Ekida fled/left Yezuha, one of the major hubs of the Engwithan empire. As such, it is a possibility that Engwithan and Ekidan are linked : either they were already separated in Yezuha and drifted further apart once the Ekida left or they all spoke some variation of Engwithan that developped independently from each other after the Ekida left. It is also a possibility that, though they lived on the same continent, they spoke entirely unrelated languages. If that is the case, it would still be probable to see loanwords or event shared structures depending on the level at which the two people were connected.
Either way, to truly know wheter these languages are related or not, we're going to have to take a look at both. Thankfully, Engwithan (and the related Glanfanthan) are somewhat documented on the Avowed wiki, so we'll just have to do the comparison work. There's still only a very few words.



Now, let's leave the realm of Eora and look at the real world info we have about these three languages. According to the wiki, Engwithan and Glanfanthan are inspired by Celtic languages: Engwithan is based on cornish (or Gaulish) and Glanfanthan is based on Irish. If we are to assume (which is unlikely considering the previous analysis) that these two languages are related to Ekidan, Ekidan would be based upon another celtic languages. Though it requires more research, there arent many celtic languages yet alive to chose from today. Among them are Scottish Gaelic, Breton and, to a much lesser extent, Manx. If we opperate under the assumption, these three are the most likely targets for inspirations for Ekidan.
It seems, however, unlikely that that Engwithan and Ekidan are in any way related, as an cursory analysis of their morphology reveals very little similarity. It would, however, be foolish to completely discredit this theory as we do not know how long these languages have been separated. We do know that their culture is nothing like that of the Rauataian and the Ixamitl, their art being unique as well as their atheism. This geographic, temporal and cultural distance could be an argument in favor of both thoeries, though so, again, more research is needed.

State of the art phonology of Ekidan
Preamble:
I work under the assumption that every letter represents a phoneme.

Consonnants: