The Symbolic Process and its Integration in Children
Chapter 4: First Words
John Fordyce Markey
IT is the third period, the basis for which has been laid by babbling, cooing and random articulation, when the child begins to repeat or attempt to repeat the words of others which concerns us. With this parrot stage already established, the problem is to determine when the true language symbol is integrated.
The data regarding speech development is in the main observational and thus lacks the control which the laboratory gives. The so-called first word is generally listed when a definite vocalization is used to designate a definite act or object. It ought to be clear immediately that such a first word may be wholly on the basis of the conditioned responses as shown in Fig. A, minus the interchange, without involving a true language habit. The attempt of a fond parent to determine the first word used " with meaning " is to be taken with some reservation.
Bateman (1917) has summarized 35 of these "first words" obtained by 28 observers (Table I). The largest group falls at 10 months; 42.85% are in the 10-11 month group, and 74.28% said the first word under 12 months (Table II). The acts or objects indicated by these first words are
|14 persons (including two cases where either dady or dog were first)||40.0%|
|18 acts associated with persons (including 3 cases of da for there)||51.4%|
|1 animal (adding the two cases cited above)||2.9%|
|2 mechanical sounds||5.7%|
Source Bateman, 1917
|Authority||Sex||Time, months||Word||Meaning of word||Time of Next word, months|
|English Speaking Children —|
|Bateman 1914||F||10.5||Hello||Same time|
|Bateman 1915||F||10.0||Daddy||Same time|
|Bateman 1916||F||11.0||Bye-bye||Same time|
|Bohn 1914||F||9.0||Daddy||Same time|
|Boyd 1914||F||11.0||Dada or dog||Same time|
|Grant 1915||F||12.0||Bye-bye ?||Same time|
|Hall 1896-1897||M||8.0||Bye-bye ?||9|
|Jegi 1900-1901||F||12.0||Mamma||Same time|
|Major 1906||M||14.0||Baby||Same time|
|Mickens 1897-1898||F||11.0||Mamma||Same time|
|Moore 1886||M||10||Papa or Mamma||—|
|Moyer 1911||F||9||Hark||Same time|
|Pelsma 1910||F||10||Daddy or dog||Same time|
|Pollock 1878||F||13||Dada||Same time|
|Shinn 1905||F||10||Da||There||Same time|
Source Bateman, 1917
|Authority||Sex||Time, months||Word||Meaning of word||Time of Next word, months|
|German-speaking Children —|
|Ament 1899||F||11.5||Mam mam||Food||15|
|Linder 1882||F||9.5||Papa||Father||Same time|
|Schneider 1903||F||10||Take take||Dancing||111.5|
|Stern 1907||F||10.5||Didda||Tic toc||11|
|Strumpell 1880||F||10.5||Ssi-ssi||Tea machine||10.5|
|Togel 1905||M||14||O (hoch)||Up||15|
|Deville 1890||F||13.5||Papa||Father||Same time|
|Taine 1876||F||10||Wawa||Dog||Same time|
|Gheorgov 1905||M||13.5||Dza ||There ||Later|
|Gheorgov 1905||M||15||Boc||Up (?)||Later|
|Altuszewski 1897||M||13||Papa||Food||Same Time|
| dza = daj, gib|
|Time of Using First Word
Source Bateman 1917
Adding all those pertaining to persons, including the two cases where dog or dady was also used, they comprise 91.4 % with only three words falling outside this category -dog, didda for tic toc, ssi-ssi for tea-machine. This is one indication of the role of persons and their behaviour in these so-called first words. No doubt parents have facilitated the application to themselves of the sounds which the infant can more readily utter. The prevalence in so many languages of some form of the root ma-ma for mother, pa-pa for father, and other similar derivations strongly suggest this (Buckman 1897, Jespersen 1923, pp. 154-160). The parents and adults are apparently better imitators than the infant.
It is quite a burden upon credulity to accept these words as real language habits. Such words as da for there, Jidda for tic toe, ssi-ssi for a tea-machine can scarcely be anything but direct responses to stimuli, and not the production of absent stimuli. For instance, it seems quite naive seriously to maintain that a nine-months-old baby said " hark " with symbolic reference. It is highly questionable whether any out of the whole list can be put in any other category than the parrot talk of infants, not beyond the conditioning shown in Fig. A, minus the interchange. The criterion used for their selection is usually this. The definite association of a name with an object does not of itself indicate real language. However, this does not diminish the fact that these words are associated with persons, which, of course, is to be expected in view of the close association of people with the learning of language. The role of persons is further indicated by the fact that if the names for the parents are not the first,
(42) they are among the first words. The acquirement of proper names is also significant. But it is a rather curious fact that quite a number of observers exclude proper names in their computations of vocabularies. This is remarkable, even ridiculous, in view of the fact of the association of persons in a child's life. Mrs. K. C. Moore found in her investigation that proper names comprised her child's total vocabulary at one year, and also played an important part thereafter (1896). Among the first words, usually occurring very early, is also a word to designate the child itself, generally the word " baby." As the time of the beginning of true symbolic behaviour is obviously not given in these so-called first words, the question must be deferred until more data are considered.
In the above list of first words even those which are not connected with acting organisms are associated with action in the form of sound. The action content of words will be further indicated in the next chapter.
After the child has once begun to acquire symbolic designations, the process soon becomes rapid and pronounced. The cases upon which a vocabulary count has been kept  are generally so lacking in any basis of random sampling that figures given for the total vocabulary at different ages cannot be said to represent the population as a whole for these ages. Consequently, no general curve of learning can be drawn from the figures ; at best they are only approximate. The comparability of one year with another in such data is also questionable, due to varying factors. There is a tendency to notice and report large vocabularies and to overlook smaller ones. Also, probably due to greater convenience, there is apparently a larger number of observations on children in the more educated or well-to-do families, which would tend to select in favour of larger vocabularies. Of course, beginning with the school period, selection obviously enters tending to produce a larger vocabulary figure, especially at certain
( 43) ages The averages in Table III, Section A, are given for what they may be worth, and are only suggestive. The vocabularies averaged by Mrs. Nice are taken from different sources, but she makes a practice of deflating for plurals, verb inflections, and the like according to Bateman's rules, so that the figures are thus more representative and more comparable. However, the rapidity of increase is evidently too steep for the general population
|Age||Authority||No. of Cases||Average|
|2||Nice (1918)||25||508 Probably unduly large|
|5||Nice (1917)||2||4225 Very large; probably nearer 2500|
Average size of vocabularies 
|Age||Number of cases||Number of Words|
|1 (1926) — figures for less than two years, from list from mothers|
|TABLE III — continued.
Kirkpatricks test 
|Graduate Students||20000 to 100000 est|
|1 (1907) --Based upon sampling from dictionary, estimates obtained from among 2,000 answers of school children mainly in Massachusetts cities, and upon Bryn Mawr, Smith, Columbia, Brown, Pratt Institute.|
|Grade||Number of Cases||Average|
|1 (1918) —From 2,000 pupils in 16 schools in Wis., Mo., Colo., from towns of 1200-30000|
for these ages. Observations have it that girls begin to talk somewhat earlier than boys and keep ahead of them in- later development as well (Stern, 1924, p. 143). If this is true, Bateman's average for one-year-olds is probably high, as about two-thirds of them are girls. Gesell (1925, p. 217) says that the median 12-months-old child has at least three or four words.
(45) Nice (1918) gives as standards of early speech development the following: the first word by 15 months, 200 words and the sentence by 2 years, Goo words and all parts of speech by 3 years. Besides the figures obtained by directly counting vocabularies, those obtained by vocabulary tests are also significant and give probably a more accurate basis for purposes of comparison on account of the fact that the tests have been standardized and used upon larger numbers of children. Some of the figures obtained by vocabulary tests arc also given in "Table III, Section B. Whipple's (1908) study in general corroborates the figures of Kirkpatrick and Brandenburg. Terman's figures (1912, not included here) are lower, but are perhaps not so comparable with the other figures, due to the method of selecting the test words. They were selected from a smaller total (18,000 vest pocket words) and with reference to the more commonly used words. It is possible that although his test is a good battery test for intelligence scoring, it may still not test the total vocabulary range.
There is a great amount of individual and group variation. For instance, Paula Lambroso (Chamberlain 1904) `found in a study that 50 children belonging to the well-todo and educated families had much larger vocabularies than 100 children of poor families. Both in the precocity with which they interpreted the words and the exactness reached the children of the former families exceeded those of the poor in the proportion of 2 : 1. She concludes that the question is not one of defect of intelligence, but of differing situations. Drevers (1919) in a study of Edinburgh Free Kindergarten children, coming from one of the poorest localities in Edinburgh, gives the following averages taken from a sample which he tried to make representative
|4||children about||3||years averaged||376 words|
|5|| ||4|| ||451 |
|12|| ||5||580 |
Some of these Kindergarten cases, according to Drevers, were not mentally normal. Due to the small amount of
(46) time spent in observation in at least four cases, these averages for the children are a little low, the three and four year averages probably being the most unreliable. Based upon similar methods of observation and other data, he estimates as a normal vocabulary from 1,000 to 2,000 words for a child of five years in good social circumstances. Allowance for an underestimation of the poorer vocabularies would still leave a substantial difference. Alice Decoeudres has worked out a language measure for young children and has used it for the ages 22 to 72. The results show the language acquirement of the working-class children to be below that of the children of educated classes. Stern (1926, p. 176) calculates from her figures for the two groups that, on the average, the children of the educated classes are eight months ahead of the working-class children of the same age (2 1/2 -- 7 1/2). Social influences obviously create individual differences and there are also structural characteristics both inherited and acquired which may influence speech behaviour.
The number of words in common among the children of even the same family may be relatively small. Gale (1900) found among three children of the same family, each vocabulary taken at 22 years of age, that out of 2,170 words, only 489, or 22.5'%, were in common. The two younger children had only 16% in common. Holdren's two children had 40% in common (Gale igoo). Drevers found among the above Free Kindergarten children that the common words were as follows (I have computed and include only percentages based upon the total vocabulary, 2,116 words for the 21 children)
|19||subjects had||2 %||of the total words in common|
|15|| ||7 %|| |
|10|| ||16 %|| |
|3|| ||52 %|| |
Thirty-three per cent. of the total words appeared in only one vocabulary, not being in common. Drevers considers that the fundamental words in the vocabulary of
(47) this type of child may be reckoned under the conditions of the investigation, as all words given by 75% of the children. The 15 cases would probably be a high 75% due to the four short observations. These had Only 7%, or 148 words in common out of a total of 2,116 words. These facts indicate some of the wide diversities which may enter into the individual learning curves, of which some examples are given in Chapter VI.