would have significantly reduced. As discussed by Crain and Lillo-Martin (1999), from post-stage V, it is increasingly difficult to explain a child’s language development in terms of different stages. Some would argue, for example, Lenneberg (1967), that this is due to the lateralisation of the brain, although others would argue against this theory.
There were many major diaries published within this era, including those of
H. Taine (1876). Others include those of Stern and Stern (1907) and Preyer (1887). Preyer’s work was probably the most extensive, as suggested by Ingram (1989). These diarists formed part of the European interest in child language acquisition. It should also be noted that there was also much interest in North America, which would appear to have been influenced by G Stanley Hall of Clark University. It could be seen that Hall followed in the footsteps of Taine by being the influence for others to write, as suggested by Ingram (1989). However, it would appear that Hall never carried out any such studies himself.
Diary studies have many advantages in their methodology. Firstly, there is the absence of observer’s paradox. This is a foreseeable problem within conflicting studies however, diary studies are generally parental, which encourages a natural environment for the child and can also ‘distinguish common behaviours from idiosyncrasies’ (). It is also a fact that a parent would have been more likely to observe significant features, as the child would have spent more time with the parent. This would have ensured that relevant facts were noted upon discovery, which would have allowed greater accuracy of the study. It would appear that linguists might have found diary studies to be comprehensive and easy to absorb due to the mass amount of information that could be assimilated in such little time. Diary studies might also have provoked the thoughts of many linguists, causing them to follow up certain data with further research.
As with all methods of study there must be disadvantages, and with diary studies these are in abundance. It is a fact that any parent can be doting towards their child, which could suggest that the study could be inaccurate and therefore, not a true cross-section of the particular developmental stage. Without accompanying evidence,
Kim Lucas 3
for example, audio/visual evidence, it is also possible that the findings may not be a true representation of the child being studied. Another major disadvantage of diary studies lies within the diarist. As previously mentioned, the major studies were parental, some carried out by linguists and probably therefore, by those of high socio-economic status. This again, results in the diary studies not being of a sufficient cross-section as it is probable that parents of such status would have been educated. This could mean that the children who were studied may have been exposed to a more complex language. Many diary studies were also carried out by those with no specific linguistic knowledge and therefore, as Ingram (1989) suggests, the observer may not be qualified to undertake such research. This could mean that significant features of the child’s language development, for example, the phonetic features, are omitted from the data, which again, could result in the inaccuracy of the study.
Diary studies can prove to be subjective and also lacking in the depth of study. Often, facts are not sufficient and require surrounding information regarding the context. There is no doubt that diary studies provided a wealth of information but, as suggested by McCarthy (1954),
Although this wealth of observational material provided stimulating and suggestive for later research workers, it has little scientific merit, for each of the studies employed a different method; the observations were for the most part conducted on single children who were usually either precocious or markedly retarded in their language development; the records were made under varying conditions; and most of the studies were subject to the unreliability of parents’ reports
(McCarthy 1954, p.494)
1926 saw the emergence of large sample studies, which meant that the previous diary studies became dormant. Diary studies did by no means cease to exist at this time, Leopold conducted a study 1939-1949 (files.lancs.ac.uk/chimp/langac/LECTURE2/2diary.htm). A more recent diary study can be dated to Weir, 1962. However, at around the time of 1926, behaviourism was having an impact on those associated with child development. This incorporated the work of Pavlov and Skinner who promoted the ideas of classical conditioning and operant
Kim Lucas 4
conditioning, as well as other protagonists of the movement such as Thorndike and John Watson. Such behaviourists believed that diary studies were incomparable; they wanted to generalise and talk about ‘norms’ and ‘explain acquisition by assigning to the child very little innate behaviour’ (Ingram 1989, p. 18). As with diary studies, the title is self-explanatory and deals with the amount of children sampled, which is larger than that of a diary study, which focuses on one child. The methodology of large sample studies more variables such as children from various socio-economic groups and also had equal gender difference. It is thought that if there is sufficient number of children under observation then a more typical behaviour will be observed. Each of the children was studied in the same systematic manner, which again, would determine some kind of typical behaviour.
Unlike dairy studies, large sample studies had the ability to identify children who were especially talented or children who were obtuse. The data collected in large sample studies was presented as measurements and as noted by Ingram (1989), consisted of mainly percentages and proportions. One advantage of this is that linguists and others interested in child language acquisition were able to extract and analyse information relevant to their particular cause. One major advantage that these studies had was the fact that the same researcher observed all children, making the study even more systematic. As the researchers were all qualified linguists, they were able to acquire the relevant details, such as phonetical differences. This also meant that the linguists did not have to worry about the inaccuracy of the studies caused by bias parents. This helped overall to make the studies more scientific.
However, there are many flaws in the methodology of large sample studies. This particular research was not inclusive of many areas of child language acquisition. Table 1.0 on the following page shows the characteristics of some of the major large sample studies that were carried out within that period.
Kim Lucas 5
Table 1.0 Some general characteristics of some of the major large sample studies conducted between 1926 and 1957
Study Simple Characteristics Topic
Smith (1926 124 children between 2 and 5; Lengths of sentences and general
One-hour conversations aspects of sentence development
McCarthy (1930) 140 children between 1;6 and 4;6 Length of sentences and general
50 sentences each aspects of sentence development
Day (1932) 160 children between 2;0 and 5;0 Study of language in twins
50 sentences each
Fisher (1934) 72 children between 1;6 and 4;6 Study of gifted children
three-hour samples
Davis (1937) 173 singletons, 166 twins, all Comparison of twins with
Between 5;6 and 6;6 singletons
50 sentences each
Young (1941) 74 children between 2;6 and 5;5 Comparison of lower- and
Six hours of conversation middle class children
Templin (1957) 430 children between 3;0 and 8;0 Length of sentence and general
50 sentences each aspects of sentence development
(Ingram 1989, p. 14)
The lack of significant features is evident. Researchers looked at the sentence length, speech sounds and vocabulary and did not account for the rules that surround language acquisition. The findings that occurred from large sample studies were all in written form and as diary studies, there was no audio/visual evidence to support the data. This raises issues again surrounding the accuracy of the studies. Is it possible for one person to accumulate so much data in so little time and ensure the accuracy? It is probable that within the area of speech sounds for example, it would be a challenge to make detailed transcriptions without the aid of audio equipment. An unnatural environment could also affect the outcome of large sample studies, as it is likely that ‘observers paradox’ would have played a role, that is, one or more of the children could have been affected by the presence of a stranger. If this is true of a large sample of the children studied, then again, the outcome may not be a true reflection of the capabilities within a specific age group.
It would be arduous to favour one of the above methodologies over the other as both types of study show clear strengths and weaknesses. As previously
Kim Lucas 6
discussed, the late 1950’s saw the emergence of longitudinal sampling, which involved the work of many linguists including that of Braine (1963, Miller and Ervin (1964), Bloom (1970) and Brown (1973). A more recent longitudinal study was carried out by linguist Clare Painter, published by Cassell, 1999, as noted in the Language Journal of the Linguistic Society of America (September 2003). It could be said that longitudinal studies provide studies of child language with an overall equilibrium by withdrawing the positive features from the methodologies of diary studies and large sample studies. It is true to say that both diary studies and large sample studies have been valid in their own right and have provided linguists with a wealth of invaluable information. The methodologies used may not have been fully comprehensive but have provided further studies with an invaluable benchmark in which to acquire knowledge of child language acquisition.
Kim Lucas 7
Bibliography
Craine, C and Lillo-Martin, D. (1999) An Introduction to Linguistic Theory and Language Acquisition. Oxford: Blackwell Publishers.
Ingram, D. (1989) First Language Acquisition: Method, Description and Explanation. Cambridge: Cambridge University Press.
Pinker, S. (1994) The Language Instinct. London: Penguin.
Painter, C. (2003) ‘Learning through language in early childhood’ Language Journal of the Linguistic Society of America 79, pp.663-664.
Lancaster University
(2 December, 2004)