The knowledge of the Iberian language is very limited, but some aspects are known.

PHONOLOGY.

    According to the written documentation we detect the following phonemes:

    five vowels: /a/, /e/, /i/, /o/, /u/ plus one or more nasalized vowel (probably allophonic)
    one nasal: /n/ one lateral: /l/
    two rhotics : normal trill and ¿retroflex or uvular? rhotic
    two sibilants: normal sibilant and ¿affricate or palatal? sibilant
    five occlusives: /b/, /t/, /d/, /k/ and /g/ (plus allophones of /b/: /p/ and /m/)

    With regard to its position in the word, they are worth noting the limitations of both trills and the lateral. No trill can be found beginning a word or a lexeme (whether it does in suffixes is not clear, but in any case exceptionally). An accented trill is the usual sign in medial plosive positions before occlusive, though both l and r can be found in that position. In absolute final position is very frequent unaccented r (which is rare in medial positions), and whereas in that position accented r does can be found, l almost never appears as final (the only clearly attested cases are the Personal Name bases sekel and tetel and even so it isn't the only spelling).

    There are also some phonetic group phenomena attested. For instance, in a compound boundary the regressive assimilations of trill and lateral before sonorants (/l/ and /n/) are usual: so biur' plus nius usually will be spelled simply biunius; and (specially in latinized personal names) the /n/ ending the first base and the /b/ at the beginning of the second one make a nasal assimilation turning out the nasal allophone of /b/, /m/. That is, /n/+/b/ > /m/. In latinized names M, in native script sometimes only b, but also not denoting any assimilation as nb.

    Concerning to the sibilant phonemes it turns out that after /l/, /n/, and vowel plus /i/ it is only accepted s, being accented s absolutely exceptional in those positions.

    With regard to the problem of the voiced and voiceless occlusives it is well known that in the Iberian personal names written in Latin inscriptions a previous sibilant devoices the following occlusive (something evident in the case of b/p), but there is a fact usually ignored: no example is known concerning to the sign s.

    The Catalan Levantine Iberian inscriptions (those which differentiate "voiced" and "voiceless" occlusives) show also a voicing progressive assimilation after /l/ and /n/ (no following voiceless occlusive is attested).

WORD ORDER.

   I have proposed the following word order description for the Iberian language:

      OV ??     Po     GenN     NA??     DemN?

   That is: maybe the Verb after the Object; Postpositional language (maybe prefixing was limited to the verbal morphology); Genitive before the qualified that qualifies; maybe the Name before the Adjective; and probably the Demostrative before its Name.

NOMINAL STEM FORMATION

    Although there must be "athematic" nouns (bare root nouns), there are attested two methods to create new nominal stems from (at least) other nominal roots. The first is the two parts compounding and we will deal with it in the next paragraph), the other is the use of derivational suffixes. It is not always easy to determine whether a brief "suffix" is truly a suffix, instead of a short stem in a compound, but probable derivational suffixes are -e, -i, -o, -no, and -ta. They are usual in place names as ars-e, iltur-o, bar'ke-no , and iltir'-ta; whereas we find -i in personal names as sakar'-i. In fact, compounding is the most frequent construction for personal names, but seemingly less important than derivational suffixes among place names.

   Probably it should be also classified as a derivational suffix the "primary" suffix -k-, which seems to have been a pluralizing suffix.

NOMINAL COMPOUNDS AND PERSONAL NAMES.

    This is the most well known aspect of the Iberian language. There is a class of nominal compounds which is the most usual way for constructing personal names, although not every compound of this kind was a personal name. There have been identified about one hundred and forty recurrent bases used for that purpose that have been extrapolated from certain Iberian personal names; attested in Latin inscriptions and texts (especially the Ascoli bronze plate), in ownership's marks, and as the variable term of grave inscriptions.

    So we have the usual elements atin,balke,bas',beles',bilos,iltir',iltur and tibas' and the personal names: atin-beles',beles'-bas',balke-atin,iltu-beles',bilos-tibas',iltir'-bas',iltir'-atin, and so on. See the major bases of the personal name compounds with examples. Seldom, there is a kind of infix between the two elements, identified are -i-, -ke- (and maybe both as -i-ke-), and -bo-.

    This brief theoretical description is enough for an epigraphical approach to the inscriptions, but from a linguistic viewpoint there are a lot of subtleties (see the bibliography).

PERSONAL NAMES SUFFIXES

    Iberian language is postpositional and thanks to the accurate knowledge on personal names it has been easy to identify nominal suffixes, some of them with an approximate meaning.

   -ar and -en: "of" / "to"? They have an equivalent use showing the owner of an object and the addressee of a document. So they have a Genitive value, maybe with some Dative functions.

    -sken/-ken: "of"? This suffix is almost only attested on coin inscriptions, where it suffixes tribal names and probably also some town names (although it isn't the usual construction with town names on coin inscriptions). The comparison with the typical ancient coin inscriptions and its "collective" use has suggested its interpretation as Genitive Plural, which is commonly accepted. It uses to be analysed as -k-en: Plural suffix + Genitive Suffix; whereas for the remaining -s- De Hoz has suggested it to be a denominative suffix for deriving ethnical adjectives.

   -ka : "to(wards)": Suffixing persons to whom it must be delivered merchandise or money, but with a too restrictive use for considering it a pure Dative; some kind of "Destinative" or Alative?.

   -te: "from" / "by". Showing the Agent of some actions, and the sender of a document. Value of Agentive and/or Ablative.

   Other known suffixes are -ku (maybe akin to -te), -ke, -u, and -e; but their function is unknown. Sometimes before the suffixes there is a seemingly unmeaningful (?) -i- (so -ika, -ite or -iu).

   Seemingly many of these suffixes could be preceded by the presumed derivational suffix, the "primary" suffix -k-. So the forms -ki-ka, -ki-te, besides of -k-en, would be a plural form (or the only plural form?) of -ka, -te,and -en, and so on.

VERBAL FORMS

    Only two are well known, although their morphological category remains unclear. The form ekiar / ekien shows the maker of the object, equivalent to FECIT. And, as stated by Untermann, the form eban / ebanen shows the builder of a monument, especially funeral, equivalent to CURAVIT (the alternative hypothesis that it means "son", advocated by Velaza, is feeble).

   There are also some (seemingly) very complex "words" for which the most obvious interpretation is that they are verbs. Even if reality uses not to be as simple as that, specially among agglutinative languages, some of these segments do seem verbs: for instance the ubiquitous forms with -(e)r'oke-. I have suggested as "typical" "verbal" "prefixes" (please, note the quotes) e-, ba- and se-.

OTHERS

    There's data about the meaning and function of some other "words". The most worth noting are the following: determiners seem to be ban ("this"?) and ir'-; iltir' and iltur have a meaning similar to "town"; ar's something similar to "castle"/"fortress"; seltar is the name of the tomb monument or simply "stela"; s'alir a monetary and / or weight unity; abiner is equivalent to "slave" in a bilingual inscription; baikar appears on ritual containers for libations. Some other words could be added to this brief list, but with a more imprecise interpretation.

    A word very well attested in Iberian (especially in ownership's marks) is m'i (alternative spelling nai). Its use clearly suggests a meaning as "this is" (less probably "I am"), but whether it is a verb, a pronoun or a deictic particle remains unclear.