In the 19th century, Francois Gouin went to Hamburg to learn German. Based on his experience as a Latin teacher, he thought the best way to do this would be memorize a German grammar book and a table of its 248 irregular verbs. However, when he went to the academy to test his new language skills, he was disappointed to find out that he could not understand anything. Trying again, he similarly memorized the 800 root words of the language as well as re-memorizing the grammar and verb forms. However, the results were the same. During this time, he had isolated himself from people around him, so he tried to learn by listening, imitating and conversing with the Germans around him, but found that his carefully constructed sentences often caused native German speakers to laugh. Again he tried a more classical approach, translation, and even memorizing the entire dictionary but had no better luck.
When he returned home, he found that his three-year-old nephew had learned to speak French. He noticed the boy was very curious and upon his first visit to a mill, he wanted to see everything and be told the name of everything. After digesting the experience silently, he then reenacted his experiences in play, talking about what he learned to whoever would listen or to himself. Gouin decided that language learning was a matter of transforming perceptions into conceptions, using language to represent what one experiences. Language is not an arbitrary set of conventions but a way of thinking and representing the world to oneself. It is not a conditioning process, but one in which the learner actively organizes his perceptions into linguistics concepts.
The series method is a variety of the direct method in that experiences are directly connected to the target language. Gouin felt that such direct "translation" of experience into words, makes for a "living language". Gouin also noticed that children organize concepts in succession of time, relating a sequence of concepts in the same order. Gouin suggested that students learn a language more quickly and retain it better if it is presented through a chronological sequence of events. Students learn sentences based on an action such as leaving a house in the order in which such would be performed. Gouin found that if the series of sentences are shuffled, their memorization becomes nearly impossible. For this, Gouin preceded psycholinguistic theory of the 20th century. He found that people will memorize events in a logical sequence, even if they are not presented in that order. He also discovered a second insight into memory called "incubation". Linguistic concepts take time to settle in the memory. The learner must use the new concepts frequently after presentation, either by thinking or by speaking, in order to master them. His last crucial observation was that language was learned in sentences with the verb as the most crucial component. Gouin would write a series in two columns: one with the complete sentences and the other with only the verb. With only the verb elements visible, he would have students recite the sequence of actions in full sentences of no more than twenty-five sentences. Another exercise involved having the teacher solicit a sequence of sentences by basically ask him/her what s/he would do next. While Gouin believed that language was rule-governed, he did not believe it should be explicitly taught.
His course was organized on elements of human society and the natural world. He estimated that a language could be learned with 800 to 900 hours of instruction over a series of 4000 exercises and no homework. The idea was that each of the exercises would force the student to think about the vocabulary in terms of its relationship with the natural world. While there is evidence that the method can work extremely well, it has some serious flaws. One of which is the teaching of subjective language, where the students must make judgments about what is experienced in the world (e.g. "bad" and "good") as such do not relate easily to one single common experience. However, the real weakness is that the method is entirely based on one experience of a three-year-old. Gouin did not observe the child's earlier language development such as naming (where only nouns are learned) or the role that stories have in human language development. What distinguishes the series method from the direct method is that vocabulary must be learned by translation from the native language, at least in the beginning.