Although various studies have been conducted on a variety of species this question has never been accepted as resolved. Although many studies have established tenuous and even some quite firm lines of communication with other species, their success is always refuted due to an ongoing debate about what constitutes language and at what point simple communication becomes language. Whilst the issue is still in flux, several attempts have been made down the years to create a set of criteria which can be used to qualify the differences between language and communication. The most well known set of 'rules' was laid out by Hockett, initially comprising of seven different rules when he initially unveiled it in his paper 'Animal 'Languages' and Human Language' in 1959, but after repeated amendments his final set consisted of 13 rules. This model is not an effective measure, however, as some of Hockett's criteria, particularly the later amendments, appear to be deliberate attempts to make language anthropocentric and deny it to other species. A better measure of language was postulated by Aitchison in 1983  and defines language as needing to meet the following ten requirements:
Mode of communication-vocal-auditory, tactile visual or chemical-olfactory
Arbitrariness-use of neutral symbols to denote objects etc.
Semanticity-the use of symbols to mean or refer to objects/actions
Cultural transmission-handing down the language from generation to generation
Spontaneous usage-freely initiating speech
Turn-taking-conversation is a two way process
Duality-organisation into basic sounds plus combinations or sequences of these
Displacement-referring to things not present in time or space
Structure-dependence-the patterned nature of language and the use of 'structured chunks' (e.g. word order)
Creativity (a.k.a. productivity)-the ability to produce and understand an infinite number of novel utterances
Aitchison, after analysing animal an human 'language' in terms of these ten criteria, concluded that semanticity, displacement, structure-dependence and creativity are unique to humans, and that it is therefore in terms of these four that studies should be evaluated 
I intend to evaluate five animal language studies in terms of Aitchison's criteria In this way I hope to establish to what extent in each study the tested animals meet the specifications for language, and consider if any of the specifications are simply beyond animals or if, theoretically, an animal could display them all and thus qualify, in terms at least of Aitchison's criteria, as users of actual language as opposed to communication.
Study One-Teaching Sign Language to a Chimpanzee (Gardener and Gardener, 1969)
In 1931 Kellogg and Kellogg attempted to raise a chimp they named Gua alongside their son Donald and see if the chimp could be brought up as a human. The experiment proved unsuccessful and was called off in 1932 as they began to fear that Gua's presence was inhibiting Donald's development after Donald, still not talking at 19 months, began to use Gua's food bark to ask for food  . Gua herself of course failed to speak as well, and this was the greatest obstacle in the Kellogg's attempts to raise her as a human. Chimps are, unfortunately, unable to mimic human speech beyond a few basic sounds as their vocal tracts are too different. The Gardener's decided to try working around this problem by teaching a young chimp called Washoe Ameslan, also known as American sign language or ASL. In its pure form, Ameslan is constructed from one or more gestures, known as cheremes. Each chereme consists of the arms and hands set in a certain position and/or performing a specific action. These words can then be combined to create sentences which are governed by grammatical rules which differ somewhat from those of normal English  , and so theoretically Washoe could fulfil all of Aitchison's criteria, although Washoe did use a somewhat corrupted version of it. The Gardener's acquired her when she was between 8 and 14 months old, as they needed her as young as possible to take advantage of the enhanced language acquisition displayed by young chimps. The Gardeners proceeded to teach Washoe ASL, mostly using imitative learning and instrumental conditioning but sometimes physically moulding Washoe's hands into particular signs, a procedure which became increasingly unnecessary as her command of ASL increased. After the first 22 months of learning Washoe had a repertoire of 30 signs, by four she was using 132 and by the time of her death in October 2007 she could reliably use approximately 250 signs, including nouns, pronouns and verbs. The Gardeners used only ASL in her presence whenever possible to avoid confusion.
a) Mode of Communication-the Ameslan used by Washoe, whilst not strictly correct, certainly qualified as tactile-visual communication, as she saw the signs made by others and signed herself to communicate
b) Arbitrariness-according to Gross in the 1999 publication of 'Key Studies in Psychology, 'â€¦the signs of ASL are neither uniformly arbitrary nor uniformly iconic but vary in their degrees of abstraction.' The signs do not have the inherent neutrality of spoken language, but are accepted as arbitrary enough to qualify for the purposes of the criterion
c) Semanticity-although Washoe corrupted the actual signs from pure ASL, it was usually not towards the overly iconic and representative and the words themselves are all as semantic as spoken words of the English language, as though the signs were not exact their meaning was always clear.
d) Cultural Transmission-although it is possible that the animals in the other studies I investigated would have been able to culturally transmit their 'language', Washoe is the only one with whom this was tested. After her own child died, Washoe was presented with a young male chimp called Loulis, whom she quickly took on as her adopted son. Loulis initially simply copied Washoe in play, but eventually simple spontaneous requests such as 'tickle', 'drink' and 'hug' were used. Loulis learned a total of 22 words from Washoe, most by imitation, but there are a couple of reports of Washoe actively trying to teach Loulis. Once when they were being brought food, Washoe repeatedly signed 'food' to Loulis and then took Loulis' hands and physically moulded them into the 'food' configuration. 
e) Spontaneous Usage-Washoe frequently used her language spontaneously in the form of requests, and is also recorded, having never used the word toothbrush before despite being signed it repeatedly, as having walked into a bathroom and identified a toothbrush there to her trainer at the time, which is referred to by Grosse in the third edition of Key studies in Psychology (published 1999) as 'â€¦the first and one of the clearest examples of Washoe apparently naming an object or event for no obvious motive other than communication.'
f) Turn-taking-much of Washoe's communication consisted of questions. She was frequently asked to identify things as part of her training and asked questions herself when the need arose. Her language seems, from my research, to primarily consist of questions and requests and doesn't really stretch far enough to accommodate musings and abstract thought, so her conversation seemed to be defined by turn-taking in its very nature
g) Duality-Washoe seems quite capable of combining words where necessary. When people failed to comply with her requests fairly quickly (e.g. Gimme tickle), she added 'hurry', 'please' or both as emphasisers.
h) Displacement-Washoe often asked for absent objects, notably when the Gardeners came back from hospital where her first child had just dies and she signed 'baby', to which Gardener signed 'baby gone, baby finished', which drove Washoe to a state best described as grief. 
i) Structure Dependence-although some signs were always combined in the same order (baby mine, tickle me) Washoe didn't always seem to care about word order. Aitchison (1983) suggests four possible reasons for this: 1) the Gardeners were eager to reward her for correct vocabulary and so never taught her grammar, 2) signs are harder to order than words (some deaf adults have trouble with sign order), 3) lacking order was only an intermediate stage in the learning process and/or 4)she did not and could not understand the patterned nature of language. 
j) Creativity-even from very early stages Washoe showed an unprompted urge to combine signs. She insisted on calling her toilet 'dirty good' even though her trainers actively encouraged her to call it 'potty chair', and similarly the fridge was 'food drink', not 'cold box'.
So based on this Washoe fills an impressive nine of the ten criteria, although her language is limited to very superficial communication.
Study Two-Project Koko (F. Patterson 1972)
Excited by a talk she heard by the Gardeners at a western US university in 1971, F. Patterson enrolled in a course to learn ASL and began to make inquiries about acquiring a baby gorilla. In 1972 Patterson began working with a gorilla born the year before called Koko. Her goal was to see if other great apes could be taught ASL, as chimps were thought to be most likely to be able to pick up language due to their strong social groupings. Initially Patterson worked in the rich yet uncontrolled environment of the zoo, where Koko constantly had people moving and talking around her whilst Patterson attempted to teach her sign language, and Patterson claims that Koko began to understand the significance of the signs within two weeks, though at that stage her signing was mostly random and out of context. The training procedures were much the same as those used for Washoe, but Koko, despite the more solitary nature of the gorilla as a species, mastered more signs than Washoe; after three years Koko knew 127 signs to Washoe's 85, and whilst Washoe died knowing about 250 signs Koko boast a vocabulary over four times that size. The same forms of operant conditioning were used, and both Koko and Washoe were capable of taking words such as 'dirty', which initially referred to their faeces, and applying them to people and events later on, displaying an understanding of the multifarious nature of words. The major advantage Koko had over Washoe is that whilst Washoe was painstakingly shown sign language to the exclusion of all else Koko has a strong grasp of spoken English. Although she is unable to speak as gorillas, like chimps, lack the necessary oesophageal architecture, Koko has demonstrated an ability to comprehend spoken English, and has developed a particular fondness for rhymes, which are impracticable in sign language. Opinion is, as ever, divided on Koko's abilities though. As Patterson and Linden said themselves: 'According to one's perspective Koko is either a dolt who has only a shaky hold on basic vocabulary, or a bright, playful, creative creature capable of quite sophisticated innovation'  For example, even in a simple rhyming game Koko expressed her answers in different ways, as shown in the following:
Hiller: Which animal rhymes with that?
Hiller: Which rhymes with big?
Koko: PIG THERE
Hiller: What rhymes with hair?
Koko: THERE (She points to the bear) 
Mode of Communication-Koko is able to understand spoken English and her own adapted version of ASL (known as Yeerkish) and can speak in Yeerkish, and so certainly qualifies as being linguistically able in this respect
Arbitrariness-Koko's sign language is classed as arbitrary in the same way as Washoe's was (q.v.)
Semanticity-again, Koko's claims to semanticity are much the same as Wahoe's
Cultural Transmission-untested, although attempts to this day are being made to mate Koko with another gorilla to see if she would teach Yeerkish to her offspring. She has owned a number of Manx cats which she has not attempted to teach, but this is probably because Koko recognises that they would be unable to learn the language
Spontaneous Usage-Koko frequently spontaneously uses language, and there are various publicly available videos showing Koko doing so, for example a video on youtube of Koko meeting Mister Rogers and asking him about his cufflinks  She also inputs her ideas into conversations around her, for example when once Patterson, talking to a friend in the same room as Koko, said 'I can't go to L.A. every month-it would kill me!', Koko suddenly signed FROWN; Patterson believes this also confirms that Koko understands the human concept of death, which is suggested by the grief she displayed when her trainers told her that her first kitten, whom she named 'All Ball', had escaped and been killed by a car. According to Patterson, Koko and her original potential mate Michael 'regularly converse[d] in sign language', and that Koko engaged ion 'personal monologues' during relaxation periods. 
Turn-taking-as with spontaneous usage, various videos can be seen showing that Koko does take part in two way conversation, although she seems to have a habit of 'butting in' if she becomes bored of what the other person is saying or if she comes up with a question. Then again, as various humans also do this in communication I believe Koko can be credited with this ability
Duality-as the conversation transcript above shows, Koko can use single words and combinations thereof, although she does not tend to create utterances of more than three or four words at a time, and often some of those will be emphasis rather than new material
Displacement-as her reference to a cat as a word that rhymes with that shows, Koko is able to refer to absent objects, but I was not able to find any more specific examples. Despite the lack of these, various sources say that Koko is quite able to refer to events displaced in time and space, and there are claims that her potential partner Michael remembered his mother being killed by poachers and attempted to tell his trainers about it, saying: "Squash meat gorilla. Mouth tooth. Cry sharp-noise loud. Bad think-trouble look-face. Cut/neck lip (girl) hole." In response to the question 'What can you tell me about your mother?' 
Structure Dependence-the general feeling about Koko, although there are both more optimistic and more damning points of view, is that she has a tenuous grasp of syntax at best, and that grammar in general eludes her, although she almost always manages to get her point across. The problem is that most of her conversation is conducted with her trainers on hand to interpret for her, and they have been accused of being rather liberal with their interpretations. On the whole, it seems that a grasp of the structure dependent nature of language is for the most part beyond Koko.
Creativity-Koko's ability for creativity seems indisputable. For example, when being taught about jewellery, when asked what a ring was she called it a 'FINGER-BRACELET'. Some of her 'creative' answers are challenged as being demonstrative of a lack of understanding on Koko's part rather than actual creativity. Take as an illustration a time when, when asked 'What's this?' about a green pig Koko responded 'GRASS PIG'. This can be construed as either Koko using metaphor or Koko not understanding the difference between the signs for grass and green.
Koko also shows nine of the ten necessary characteristics of language, but is distinctly more advanced than Washoe in the depth of language that she uses, suggesting that even within these ten criteria subdivisions need to be made to establish the level to which each is followed if we are to truly quantify the concept of 'language'
Study Three-Akeakamai and Phoenix (Herman et al. 1984)
Although the great apes are the most obvious candidates for language studies, various other species have been experimented upon, including in this case bottle-nosed dolphins. Two were experimented upon, but each was taught language through a different medium. Akeakamai was taught a system of hand signals following a set of grammatical rules, as show in appendix A; Phoenix was taught language as a series of short discrete noises generated by a computer. For the sake of brevity, I shall focus on Akeakamai's study. The language has some distinct disadvantages over the ASL variants used for the two ape studies evaluated thus far-the most important of these is that the language was entirely one way, and although tests could be made on the dolphin's comprehension she was unable to produce sentences. The major advantage is that the simplistic and generally imperative nature of the language meant that Akeakamai's comprehension could be easily tested and measured, for example 'ball hoop fetch' means 'take the hoop to the pipe' and 'hoop ball fetch' means take the ball to the hoop, and when she was first introduced to a four word sentence Akeakamai successfully applied her existing grammar rules to it and performed correctly.
Mode of Communication-due to the one way nature of the communication, Akeakamai cannot as yet be proven to use a two-way process
Arbitrariness-the symbols are arbitrary, indeed to make them otherwise would be highly inconvenient and need to involve a simulated dolphin of some description
Semanticity-each symbol was attributed a very specific meaning as demonstrated by Akeakamai's usually correct responses. Her ability to distinguish is regularly and easily tested by PENDING EMAIL RESPONSE FROM DOLPHIN INSTITUTE, WHO ARE UNRESPONSIVE 'TIL SEPTEMBER 24th
Cultural Transmission-currently untested and would be very hard to test in any case due to the purely receptive nature of the language
Spontaneous usage-once again untestable
Duality-Akeakamai's ability to comprehend instructions of varying lengths has been clearly demonstrated through testing, so arguably duality is present, although it is receptive rather than productive duality
Displacement-this can be, and has been, successfully tested. Akeakamai has all the objects in the pool but one, which the trainer specifies in a command. Initially she would then spend a couple of minutes hunting before returning to the trainer, and has now been trained to press a paddle if the object is missing, demonstrating that she understands what is wanted but does not find it present. Temporal displacement comprehension has also been tested by throwing the object in a minute or two after the command has been given, at which point she usually carries out the set task. 
Structure Dependence-this study has the most evidence for successful acknowledgement of structure dependence by the animal of any other species study. The quantifiable nature of comprehension due to the fact that the language is used solely to issue commands means that Akeakamai has been proven to be able to interpret syntax of commands and respond accordingly, achieving success rates of PENDING EMAIL RESPONSE FROM DOLPHIN INSTITUTE, WHO ARE UNRESPONSIVE 'TIL SEPTEMBER 24th
Creativity-so far Akeakamai has been able to incorporate all news signs into her vocabulary without difficulty, but creativity cannot be truly tested due once again to the nature of the language
Akeakamai can only really be said to meet four of the ten criteria, but does successfully show that at least some other species are capable of understanding the concept of grammar and use it in language. It is tempting to say that if a dolphin can achieve it an ape can too, and the apes like Koko and Washoe should be able to, but this is an unfair judgement as it could be the case that dolphins brain structure allows for the concept but it is precluded by the apes' mindsets, or it could be that dolphins are in fact more intelligent than apes but simply harder to talk to due to the difference in communication methods between anthropoids and cetaceans.