The unitary system hypothesis and separate system hypothesis are two theories that describe how language develops in children who are exposed to two different languages, most often in the home.

First, we should discuss the basic premise of each hypothesis and how it applies to bilingual children. Keep in mind that each hypothesis applies to young children developing language skills, most often under the age of four.

The unitary system hypothesis states that children learn two languages as one in the beginning. The two languages are united into one set of rules, vocabulary, and structure. Around age three or four, children begin to separate the two languages. They learn to code-switch, meaning that they can switch from one language to the other. This is a difficult skill to master and one of the reasons that early and consistent exposure to both languages is so important when raising a bilingual child.

The separate system hypothesis states that a bilingual child is able to recognize and use each language separately from very early in the language acquisition process. In this hypothesis, children naturally learn which rules, vocabulary, and language elements go with each language through exposure, use, and practice.

The common elements of both the unitary system and separate system hypotheses are that early exposure to both languages is critical for children to learn and master two languages. Language begins to emerge around age one and develops rapidly through the second and into the third year of life. Bilingual children learn two languages when hearing and practicing them consistently during this key developmental stage.

The unitary and separate systems hypotheses are two different theories about the way language is acquired by children in a bilingual household. Naturally, children who grow up hearing two languages spoken frequently will develop an understanding of each of those languages. These theories differ, however, in whether or not there is a mother tongue and if the lexicons combine during the language acquisition.

In the unitary system, the child learns one grand, combined system of both languages—the lexicons are combined, and they will speak and reason in a conglomerate of the two languages. As they grow and their understanding of language increases, the two languages will separate themselves out through use. Also in the unitary system, there is a mother tongue, a more prominent language that is more frequently used.

In the separate system, the child will learn both languages distinctly, understanding the delineation between different sets of words and languages, and there is no need for separation down the line.

These two distinct theories are in competition to determine which one better describes how children acquire language. Code-switching and mixed speech lend themselves to the unitary hypothesis, wherein the child will speak a garbled combination of the two or will switch between a main lexicon and a subgroup of foreign words while they are learning to speak and growing older.

The separate system, however, has evidence in the fact that children naturally can choose to speak in one dialect or the other, and there is no clear age at which the "switch" occurs.

First, let's describe the two hypotheses and then analyze for a conclusion.

The Unitary System Hypothesis is quite interesting. It stems from the continuous debate on whether dual-language babies acquire one or two languages at the same time. The essential question posted in the debate is whether the babies realize that they are speaking two languages, or whether their brain acquire both languages as one same unit which they will begin to differentiate at around age three when their speech formation begins. In all, the USH basically supports the claim that children learn ONE language first, and the second later at a different rate, and NOT simultaneously.

The argument in favor of the USH is that studies made with children of dual language homes, at age two and three, show evidence of code-switching. Code-switching suggests that the child knows both languages (switching) but uses one language more than the other. Other studies suggest that the reason for the rate of usage of L1 and L2 is entirely dependant on input, and that such input is found to be imbalanced in most families. This means one thing: that USH's claim of the acquisition of one language at a time depends on how much the parents expose the children to the languages; one lexicon, one grammar. The implication for learners is that the parents must be consistent in speaking both languages at the same rate at home, or they will find the child opting for the common language of the home and then code-switching with the second one.

The Separate Systems Hypothesis (SSH) is the opposite to USH. In this hypothesis, the argument is that the child will build two separate lexicons, and two separate grammars for each language. The SSH and USH look quite similar in that, at some point, the child will indeed use both languages in their particular forms differently. If consistent input is given, the child will eventually develop two completely languages correctly. They key is input, consistency, and appropriate usage of both language groups.

Remember that L2 research is based on theory, and not on scientific discovery that could lead us to a factual source of language acquisition. Hence, both theories do intermingle, yet, they both agree in that parents must take their language usage at home seriously and not tend to favor one language over another. This is the only way to succeed in an L1/L2 home