Kaisa Launonen, Ph.D. Augmentative and Alternative Communication: European Perspectives Stephen von Tetzchner and Mogens Hygum Jensen (Eds.) Chapter 13, p. 213-31

Reproduced with the permission of the editor Whurr Publishers ©1996. All rights reserved

Method
Subjects
The project consisted of two parts. In the first part, the development of communication of 29 children was followed. The second part was a controlled study in which a subgroup consisting of 12 of the 29 children was compared with a control group.
   Intervention group. There were 36 families in the first ten family groups of the intervention programme. Of these families, 29 were included in the intervention group. Seven families were left out, either because they did not follow the programme regularly or the children had severe epilepsy which seriously affected their development. There were 17 girls and 12 boys in the intervention group. All came from urban families of which 27 were Finnish-speaking, and two were bilingual (Finnish/Swedish and Finnish/English; in the latter case, the language used with the child at home was Finnish. Finland is an officially bilingual country, and the Swedish-speaking minority in the Helsinki area is 7.1 per cent).
   Research group. The twelve oldest children of the intervention group, six boys and six girls, made up the research group. All of them had trisomy 21 and no other disability which could have had an effect on the intervention. Five of the children had a congenital heart defect which was surgically corrected at an early age. Seven of the children were first-borns. By the age of five, two were single children while the others had one or two siblings. All came from Finnish-speaking families and attended day-care outside their home; seven children from the age of one, four children from the age of three, and one from 4½ years. For most of the children, individual speech therapy started soon after the programme was terminated at the age of three. For one child, speech therapy continued to the age of four and for another child to age five years. For the rest, it went on at least until school age, which for disabled children in Finland is six years (for other children seven years).
   Control group. For ethical reasons, after it was initiated, the intervention programme has been offered to all families living in Helsinki area with a new-born child with Down syndrome. This caused some problems for forming a contemporary control group which did not participate in the programme. The control group therefore comprised the 12 oldest of the preceding age group of children with Down syndrome in Helsinki area, of whom all the families agreed to participate in the project. In the control group, eleven children had trisomy 21 and one had translocation trisomy. None of them had other impairments. Four children had a congenital heart defect; one of them could not have an operation. There were nine girls and three boys in the group. Three of the children were first-borns. By the age of five years, the children had 1-4 siblings. All came from urban families; eleven were Finnish-speaking and one was bilingual (Finnish/Swedish). By the age of five, the ten remaining children attended day-care outside their homes, having started between ages of one and four years.
   Services other than the early signing programme (see below) were the same for the two groups. Early intervention for the children in the control group consisted of services given to all families attending the clinic for disabled children, including the intervention group. For most families, these services included two yearly visits to the clinic where a team of professionals examined the child, consulted with the parents, and gave them advice, both orally and in writing, on how to enhance the development of the child in different skill areas. Nine children in both the research group and the control group were given physiotherapy before learning to walk without support. For the control group, individual speech therapy was initiated between 2½ and 5 years, with an average of 3½ years. For these children, speech therapy continued at least until they reached school age.

Procedures
All families with an infant with Down syndrome, living in Helsinki area, were, if willing to participate, included in the early signing programme. The programme started when the child was about six months old and ended at the age of three. The speech therapists' part in the intervention was mostly indirect. They gave advice to the parents who trained their own children. Direct contact between the child and the speech therapist was included in the programme in order to follow the development of the child and to ensure that the intervention was carried out according to the individual needs of each child and its family.
   Family groups of 2-4 families met with two speech therapists; during the first six months, groups met every second month, and, after that, once a month (see Table 13-1). In the course of the 2½ years of intervention, there were 25 sessions with each family group and at each session the families were assigned tasks and given advice for training at home. Besides the family groups, the child and the parents had individual appointments with the speech therapist every second week. If possible, the day care personnel were included in the programme. Training, however, was the responsibility of the parents.

Table 13.1: Follow-up of the programme according to the age of the child

Age in months	Stage of the programme
6-12	Daily:	Training at home
	Bi-monthly:	Family session
12-36	Daily:	Training at home
	Bi-weekly:	Speech therapist meets the child
	Monthly:	Family session Evaluations made by parents and the speech therapist
	Semi-annually:	Portage assessment Three questionnaires Video recording
36-60	Annualy:	Portage assessment Three questionnaires Video recording

   Parents were advised to have daily brief 'training sessions', in order to make the use of the signs explicit. Signing in daily situations at home was, however, emphasized. In the beginning, training consisted of a variety of shared activities, aimed at encouraging the child and the parents to take part in mutual activities. Many of these were traditional games (such as 'See saw Margery Daw', 'Round and round the garden', and 'This little piggy'). In the beginning, activities in which the focus is in the use of hands were emphasised, and in the course of the first six months, the focus shifted gradually to conventional manual signs. From the age of twelve months, the monthly programme included 6-10 new manual signs, suggestions for training at home, and a song which was sung with signs. The last ten monthly programmes consisted of games and play designed to encourage the children to actively use their different means of communication. During this period, teaching of new manual signs was related to individual needs with the help of video tapes and instructions given to parents by the speech therapists.
   The family group sessions, each lasting 45-60 minutes, usually started with singing the 'signing song' of the previous month, followed by the new manual signs and strategies. Parents were given drawings and verbal instructions for each new sign and pictures from children's books on the same theme, and advised to collect the graphic material in a book which the child could use as a picture book when together with other people. In addition, they were given oral and written suggestions for training of each new sign. The new song, with its signs, was taught to families. After going through the new programme, there was usually a brief period of 'free conversation', and the session was finished by singing the song of the month.
   Initially, parents were advised to train with one of the parents opposite the child, giving the model, and the other was sitting behind the child, hand guiding it to make the manual sign when necessary. At the same time, parents were advised to be sensitive to the child's reactions and to keep help at a minimum. After the child had learned some manual signs, imitation was emphasized as a way of learning new signs. Parents were told to speak normally when using signs and to sign the 'key words' of the spoken utterance, according to the language skills of the child.
   Even if the routines of the signing programme were structured, the purpose was not to make the parents slavishly follow the routines. They were told that learning manual signs was not a goal but a tool. One of the main themes in the programme was to reinforce the child's active behaviour and initiative from very early on. The parents were given varied information, both orally and in writing, about conditions that may enhance language development and on how to encourage the child to use its early means of communication. The systemic, interactive nature of all communication and learning was emphasized, as well as the fact that infants are competent individuals, responsible for their own part in communicative interactions. It was also pointed out that children need a range of interesting challenges which can encourage them to explore the world. The child's characteristics and communicative strengths were discussed by the parents and the speech therapists. The family was always encouraged to use its own individual style in communication.
   In addition to being an empirical basis of the project, the assessment methods and video recordings (see below) had aims closely linked to the intervention itself. Through taking part in assessments, evaluating their child's development, and watching the video tapes when at home or with the speech therapist in the clinic, parents learned to observe both their own and the child's behaviour. They learned to search for the strengths of the child and identify useful intervention strategies (cf. Ballard, 1991).

Measures
The evaluation of children in the intervention group was made every sixth month from 1 to 3 years of age. Follow-up assessments were made when the children were 4 and 5 years old (and will be made at the age of 8 years). General development was evaluated with The Portage Assessment Scale (Tiilikka and Hautamäki, 1986) which gives a profile of the child's skills in five areas: social, language, self-help, cognitive, and motor. The Portage scale is not standardized. It can, however, be used to evaluate the developmental level, and its reliability and validity have proved to be good in studies of intellectually impaired populations in Finland (Arvio, Hautamäki and Tiilikka, 1993). Scoring was adapted to the needs of the project in order to show effects of signing skills separately. Because some children mastered a skill only by using signs, children who used signing were given two values: with and without signing, mainly for the areas of social development, language and cognitive development. For self-help and motor development, signing skills as such did not make any difference. The Portage assessments of the intervention group were mostly made at the clinic by the speech therapist and one or both of the parents.
   Three questionnaires were created for assessing the development and efficiency of the child's expressive communication. One of the questionnaires was filled in by the parents and the speech therapist separately once a month, the other two were used in half-yearly evaluations. The parents usually filled in the questionnaires at home and if there were any unclear points, they were discussed during the next session or soon afterwards. In addition, parents made notes of the daily training at home. At the half-yearly assessments, the children were video recorded playing with the speech therapist and sometimes also with one of the parents. At the present time, some of the children in the intervention group has not yet reached the age of 4 years. Hence, this age level includes only 21 children.
   Except for video recording, the control group was assessed in a similar way once a year. Because the project started when the oldest children in this group were almost three years old, many of the early observations are not complete. Five children were assessed with the Portage scale at the age of one and seven children at the age of two. All twelve children were assessed at the ages of three and four years, and ten children at the age of five. One family moved away and one family did not want to take part in the project when the child was five years old. Thus, the most complete comparisons are available for the time when the programme was terminated for the research group, at the ages of three, four and five years. Although at this stage, some children did and some did not use manual signing, the principles of intervention were the same for both groups.
   The assessments of the control group were made mostly in the home by the speech therapist and one or both of the parents. The questionnaires were usually filled in together on the same occasion. Parents of the control group were not as accustomed to making evaluations of their child as were parents of the intervention group.
   Several milestones of the early development of communicative skills of the intervention group were registered: early gestures (waving 'bye-bye' and clapping hands), pointing with index finger, first manual sign, first spoken word, and the number of manual signs and spoken words yearly from 12 to 48 months. These measures were based on information given by the parents. A checklist of specific items such as waving 'bye-bye' or clapping was not provided. Instead, parents were asked to write down all the early gestures that the child was using. For this reason, this information is missing for some children. For pointing with index finger, it was required that the child pointed to an object out of its reach, not just touched an object or a picture with the finger. Determining when a child for the first time used a manual sign purposefully proved to be difficult. The first imitation of a manual sign, recognized by the parents and the speech therapist, was accepted. Determining the 'first spoken word' proved even more difficult (cf. Bates et al., 1975). Some parents reported the first word of the child approximately at the age of one, but could - maybe a year later - state that the child had yet not said its first word. If parents corrected their evaluation, the later age was used - if not, the first one. Thus, the estimates of first words are only suggestive.
   The expressive vocabularies of manual signs and spoken words were based on records made by parents. Because active use was difficult to define, parents were asked to write down all the manual signs and spoken words that the child could produce when asked to. In the later evaluations, when some children used dozens or hundreds of signs and words, parents were asked to estimate the vocabulary size. Pure imitations were not counted because some children in the intervention group could imitate signs almost indefinitely without understanding their meaning. The way that an individual family collected this information was not checked; parents of the intervention group knew that they would be asked for the information at the half-yearly assessments. Some parents asked for help from the speech therapist for these evaluations but both in the intervention group and the control group the final numbers were always the parents' estimates. In the evaluations at three and four years of age, the vocabularies of some children were so large that their parents just wrote 'hundreds'. In these cases it was decided to give them a size of 300, which for some children might be somewhat low.
   For comparing the research group and the control group, analyses of variance was made of the yearly Portage scores from one to five years of age and repeated measures of analyses of variance from three to five years of age. The number of manual signs and spoken words are compared at three and four years. By the age of five, expressive language had for many children reached such a level that it was impossible to estimate the number of individual signs and words. When estimating the number of symbols (manual signs and spoken words), parents also gave examples of first combinations of symbols that the children had used.

Results
Milestones in early communicative skills of the intervention group and the research sub-group are shown in Table 13-2. There was considerable individual variation with regard to when the children were reported to clap hands, wave 'bye-bye', and point. All the children participating in the programme started to use manual signs and the mean age for the first sign was 17 months. At the age of 1½ years, most of the children (23) used some signs, and by the age of two years, all the children used some manual signs for communication (Table 13-3). There were large individual differences with regard to both the number of signs and how actively and varied were the signs used. The increase in the number of manual signs was notable until the age of three. By that age, some children began to speak quite clearly, and by the age of four, the average number of spoken words had surpassed the number of manual signs. Seven children of the 21 children in the intervention group used more spoken words than manual signs in their communication at the age of four.

Table 13.2: Age in months at appearance of early communicative skills of children who participated in the early signing programme; intervention group (IG, N=29) and research group (RG, N=12)

		Mean	Std	Range	Median	Missing
Clapping	IG	11.4	2.6	6-16	11	12
	RG	11.0	0.6	10-12	11	7
Waving	IG	14.7	3.4	9-23	14	7
	RG	13.6	3.4	9-21	13	4
Pointing	IG	16.0	2.3	10-22	16	0
	RG	16.7	1.9	14-22	16.5	0
First sign	IG	17.0	2.2	14-22	17	0
	RG	17.8	1.8	14-21	17.5	0
First word	IG	18.8	4.6	12-30	18	0
	RG	19.3	4.8	14-30	18	0

   Table 13-3 indicates, that at ages of three and four, children of the research group were using more communicative symbols, both manual signs and spoken words, than the children in the control group. When manual signs and spoken words were counted together, the size of the average vocabulary of the research group at age three was as large as that of the control group at age four. All children in the research group used manual signs in their communication. In the control group, three children communicated by signing (at age four) and these three were among the most advanced linguistically of the control group.

Table 13.3: The number of manual signs, spoken words, and signs and words counted together (total) of children of the intervention group (IG; N=29, at 48 months N=21), research group (RG; N=12), and control group (CG; N=12). (Med = median; 'None' = number of children without any manual signs or spoken words)

		Manual signs					Spoken Words					Total
Age	Group	Mean	Std	Range	Med	None	Mean	Std	Range	Med	None	Mean	Std	Range	Med
12	IG	0	0.0	0	0	29	0.2	0.5	0-2	0	25	0.2	0.5	0-2	0
	RG	0	0.0	0	0	12	0.2	0.6	0-2	0	11	0.2	0.6	0-2	0
18	IG	7	9.8	0-40	3	6	1.3	1.7	0-6	1	13	8.3	9.9	0-41	4
	RG	0	0.0	0	0	12	0.2	0.6	0-2	0	11	0.2	0.6	0-2	0
24	IG	31.4	25.6	3-100	20	0	5.7	6.7	0-32	3.5	2	37.1	26.6	4-111	29
	RG	34.2	21.1	10-74	27	0	3.9	3.3	0-10	3.5	2	38.1	21.7	11-81	31.5
30	IG	73.3	55.1	5-200	65	0	11.9	18.3	0-100	8	1	85.1	61.6	8-220	74
	RG	74.2	42.6	15-150	80	0	7.6	5.9	0-20	7	1	81.8	43.2	19-158	87
36	IG	101.7	79.0	10-300	100	0	17.3	20.1	1-100	10	0	118.4	84.5	17-318	102
	RG	93.3	65.0	15-250	100	0	17.3	17.0	1-50	11.5	0	110.7	73.2	20-300	106.5
	CG	3.3	4.6	0-17	2.5	5	10.3	11.2	0-40	6.5	3	13.6	11.2	0-40	12
48	IG	151.0	104.1	20-350	200	0	105.0	122.0	3-300	200	0	255.9	141.1	25-500	305
	RG	108.3	69.1	20-250	90	0	128.2	131.8	3-300	50	0	236.5	141.6	30-500	220
	CG	35.7	82.8	0-300	1	6	75.8	98.2	0-300	45	1	111.4	121.8	0-355	65

   At the age of three, all except one of the research group joined two or three symbols (signs, signs and words, or words) together, two of them only occasionally. In the control group, three of the children joined two spoken words together, and two children joined two manual signs occasionally. At the age of four, four children in the research group used spoken sentences, four joined manual signs or signs and spoken words together, and four used such combinations occasionally. In the control group, seven children joined two or three spoken words together. At the age of five, six of the twelve children in the research group used spoken sentences, one used manual signs and spoken words in all three ways of combination, one joined signs and spoken words together, and four used combinations of two signs or signs and gestures. In the control group, five children used spoken sentences, one combined signs, signs and words, and words. Four out of ten remaining children in the control group did not join symbols together at the age of five (Table 13-4).

Table 13.4: Number of children in the research group (RG) and the control group (CG) who made different combinations of symbols at the ages of 36, 48, and 60 months. W = combinations of spoken words; W+S = combinations of spoken words and manual signs; S = combinations of manual signs

	36 months		48 months		60 months
	RG	CG	RG	CG	RG	CG
W		3	4	6	6	5
W, W+S				1	1
W, W+S,S	2		1		1	1
W,S	2
W+S,S	1		3
S	6	2	4		4
No combination	1	7		5		4

   On the first Portage assessment at 12 months, the two groups did not differ significantly. The children in the control group (N=5) seemed to be slightly ahead of the research group (N=11) in all areas except self-help (Table 13-5). At the age of two, the research group (N=12) was ahead of the control group (N=7) in all areas except self-help. The difference was greatest for language development with signs, the area where the intervention programme could be expected to have the most immediate effect.

Table 13.5: Comparison of children in the research group (RG) and the control group (CG) on the means of the Portage scores at 12, 24, 36, 48 and 60 months (* p<.05; ** p<.01; *** p<.001)

		RG	CG	df	F
12	months
	social	11.36	12.57	1,14	.43
	language	5.01	7.54	1,14	2.68
	self-help	9.97	8.76	1,14	.51
	cognitive	5.07	6.34	1,13	.76
	motor	6.81	7.96	1,14	.78
24	months
	social	26.50	23.50	1,17	1.28
	social + signs	27.78	23.50	1,17	2.34
	language	14.91	12.03	1,17	3.58
	language + signs	17.90	12.03	1,17	11.42	**
	self-help	21.45	22.10	1,17	.09
	cognitive	18.75	16.01	1,17	2.53
	cognitive + signs	19.88	16.01	1,17	4.15
	motor	20.60	19.14	1,17	.28
36	months
	social	42.63	37.29	1,22	2.85
	social + signs	45.50	37.29	1,22	6.54	*
	language	23.34	18.77	1,22	4.69	*
	language + signs	29.78	18.77	1,22	22.34	***
	self-help	38.93	35.26	1,22	1.44
	cognitive	30.89	23.61	1,22	23.76	***
	cognitive + signs	32.28	23.61	1,22	30.52	***
	motor	37.94	33.14	1,22	4.32	*
48	months
	social	52.10	47.81	1,22	.89
	social + signs	54.60	48.23	1,22	2.11
	language	34.41	23.11	1,22	7.44	*
	language + signs	38.34	23.86	1,22	15.32	***
	self-help	47.79	43.29	1,22	1.39
	cognitive	37.39	28.42	1,22	12.46	**
	cognitive + signs	38.60	28.69	1,22	16.11	***
	motor	45.31	40.43	1,22	2.62
60	months
	social	57.75	53.39	1,20	.88
	social + signs	59.78	53.63	1,20	1.93
	language	41.15	30.53	1,20	3.40
	language + signs	44.70	30.65	1,20	7.73	*
	self-help	52.81	49.04	1,20	.98
	cognitive	42.58	34.03	1,20	5.66	*
	cognitive + signs	43.89	34.08	1,20	8.13	**
	motor	49.44	46.88	1,20	.60

   From three to five years of age, the profiles of the Portage assessments of the research group and the control group clearly differed from each other, the children in the research group being ahead in all areas of development (Table 13-5). The difference was evident even when signing skills were not taken into consideration. The difference between the two groups was most marked at age three, when the intervention programme was completed. The research group (N=12) was significantly ahead of the control group (N=12) for all Portage areas except social development without signs and self-help. The difference was greatest in the areas of language with signs and cognitive development with and without signs. The differences between the groups had started to decrease at ages four and five (Table 13-5). However, significant differences were still found in the areas of language and cognitive skills. The repeated measures of analysis of variance revealed a significant difference between the groups in the development of language and cognitive skills, both with and without the help of signs, during the two-year follow-up after the intervention programme had been completed for the research group (Table 13-6).

Table 13.6: Differences between the research group (RG) and the control group (CG) on Portage at 36, 48 and 60 months of age (repeated measures ANOVA). (* p<.05; ** p<.01)

	df	F
Social	1,20	1.49
Social with signs	1,20	3.27
Language	1,20	4.96	*
Language with signs	1,20	12.96	**
Self-help	1,20	1.38
Cognitive	1,20	9.87	**
Cognitive with signs	1,20	13.69	**
Motor	1,20	1.94

Discussion
The results show that the early signing programme had significant immediate benefits for the children who took part and that the positive effects remained during the two-year follow-up after its completion. Compared to the control group, at three, four and five years, the children in the research group used a far wider range of communicative means and were clearly ahead both in language and general development, especially in cognitive skills.
   The main difference in early intervention between the children in the two groups was the purposefully enlarged and intensified usage of non-vocal means of communication in the research group. This was developed as far as it was individually necessary and for all the children it implied some degree of signed communication. All had a period where they obtained higher scores on the Portage scale with manual signs than without. There were, however, marked individual differences in the language development of the children in the research group also. For some children, the period where signing was dominant lasted less than a year; for others, signing was their most functional means of communication at least until the age of five.
   Availability of differentiated symbols for active use makes it possible for a child to interact with others in a way which may develop its communication even further. For example, at the age of three, having 93 manual signs made a big difference compared to having to rely on 17 spoken words only for communication, and even this was seven words more than the control group. The research group had an average of 111 symbols (signed and spoken) while the control group had 14. Manual signs made many communicative functions possible which would be out of the children's reach with nonverbal means of communication only, such as making requests, questions and comments, getting information, telling one's experiences, or even joking. Many of the parents in the early signing programme commented spontaneously during the active signing phase of their child that they had difficulties imagining how these communicative needs would have been fulfilled had not manual signing been available to the child.
   The present findings suggest that augmented language development may enhance cognitive skills and that the developments of these two areas are closely linked. As argued above, availability of differentiated symbols provided the children with opportunities for dealing with objects, people and events on a more mature level. This cognitive growth, in turn created new conditions for versatile functions of communication and language. The results support the findings of Johansson (1990), according to which preschool children with Down syndrome who had participated in an early language intervention programme had far better linguistic and cognitive competence than a non-participating control group of children with Down syndrome.
   The signing is also likely to have affected the spoken input to the children in a way that made it easier for them to obtain information and understand the utterances. Part of this was probably a natural consequence of signed communication being added to the speech, but also the interaction with the speech therapist was important for the parents' learning to adapt their communication to their child's skills. When using key-word signing, parents were likely to speak more slowly, used shorter utterances and probably put stress on words they both speak and sign (Windsor and Fristoe, 1989). A significant characteristic of simultaneous signing and speaking is also that parents had to make sure that they had visual contact with the child while talking to it. The parents also had better opportunities for observing the child's behaviour and responses and reacting in an adequate manner. For instance, they waited and gave the child time, repeated the utterance, gave extra information and continued the conversation according to the child's response. The simultaneous use of visual and auditory forms of communication may also have made it easier for the child to obtain information and thus to expand its cognitive competence.
   Even if the possible strength of visual processing of persons with Down syndrome is not taken into consideration, the visual-motor character of signs may be of significance, especially the possibility of adapting the speed of production when signs are taught to a child. Contrary to saying words, manually signing can be slowed down, sometimes even stopped, without loss of intelligibility. Signs can also be taught through hand guidance. Moreover, if visual skills are better than auditory skills for individuals with Down syndrome, it is likely that visual signs will catch their attention more easily than spoken words.
   The results indicate that manual signing in early communication intervention, apart from advancing language and general development, also enhanced the speech development of the research group. This is evident in the Portage language scores without signing where the research group was ahead of the control group on all assessments from three through five years. Also the reported number of spoken words was higher for the research group at the ages of three and four.
   Most of the children who participated in the programme started joining two or three manual signs together. The most common first combinations were, however, the joining a manual sign and a spoken word together. According to Caselli and Volterra (1990), the capacity to use symbolic and combinatorial ability simultaneously in order to communicate, indicates that the child is passing from using a general communicative capability to managing a real linguistic system. Most of these children with Down syndrome used manual signs at this transitional stage but it seems as if the signing ability never reached a 'mature' level of verbal communication even though it fulfilled versatile functions of communication. However, it is clear that the manual signs of these children, even if used only as single-sign utterances, functioned like verbal symbols in communication in the same way as single-word utterances of normally developing children. Furthermore, children who seemed to possess greater problems with speech than with language development, started to use manual signs in sentence-like structures, as a real linguistic system.
   The findings of this study suggest that the early signing period was a bridge from the early preverbal phase of communication to the use of spoken language and that the transitional phase from preverbal to verbal communication may be enhanced by a goal-oriented use of the means of communication which may be available for the child prior to the verbal symbols. However, for children with Down syndrome conventional gestural communication is not enough for the transition. They need more intensive, more long-lasting, and, it seems, different stimulation than other children. The stage has to be intensified and modified qualitatively as in the early signing programme. One dimension of this modified and added quality was the use of manual signs which were easier for the child to attain than spoken words. Another important aspect was the interpretations parents made of their children's actions. These interpretations may have advanced the children's awareness of their actions, and, accordingly, the acquisition of shared meaning between them and their parents (cf. Smith, von Tetzchner and Michaelsen, 1988).
   One of the main effects of the programme was thus to shape the child's communicative environment to its needs and abilities. Signed communication was an essential part of this beneficial environment. However, it is possible that introducing a new way of communication to the family might disturb the natural communication between the child and its parents. It is therefore important that parents get to appreciate more generally the meaning of communication and its development. It was hypothesized that if parents were given adequate information on these matters, they would become aware of the importance of the child's own active role active role in forming a conception of the world. It is likely that this awareness was strengthened through the regular conversations with a speech therapist. It may also be assumed that the regular evaluations the parents made played a major role in the development of their communicative knowledge. Parents learned to attribute competence to the child in communication even before signing appeared, when the child was using early preverbal means of communication. In this balanced situation of interaction, the child got adequate support in its initiatives and was encouraged to be challenged in both communication and exploration. An active role from very early on may be of ultimate importance for the development of the child's image of itself as a communicating individual.
   Some of the differences between the groups may possibly be explained by general effects of early intervention. One important aspect of all early intervention programmes is the support given to parents (Cunningham, 1988; Hornby, 1991). Many parents with impaired infants feel uncertain about their parental role. They may not know what to expect of the child, generally or at a given age. When given adequate information and support they may feel more secure and relaxed and begin to gain confidence in their parental role. A more relaxed parental role may have positive effects on the interaction between the child and rest of the family and, in turn, enhance the child's opportunities for developing early communicative skills. One may question whether the general support of the parents in the control group was sufficient. It is possible that the support had more positive effects when it was provided within the more concrete frame of teaching early signing. Parents may feel more confident when they are 'doing something' for and with their child. Moreover, the value of the family groups and the support parents in the same group gave each other cannot be overestimated.
   Those who criticize the use of special programmes for early intervention of intellectually impaired children often argue that what is significant is not so much the quality of the interactions as the amount of time spent with the child (e.g. Gibson and Harris, 1988). Intellectually impaired children need opportunities to learn the same skill repeatedly and in many different situations. If parents spend more time with their child, repetition becomes possible. However, children with Down syndrome are left alone to explore their surroundings more often than normally developing children (Smith and Hagen, 1984). The parents in the early signing programme were advised to have daily 'training sessions' with their child and this may have given the children more opportunities for attaining communication and other skills. However, it is not known whether the research group parents spent more time with their children than the parents in the control group. Further, knowing that the quality of the early interactions was changed by adding signed communication, the effect of a possible difference in the time spent with the child cannot be distinguished from the effects of other factors.
   Because the groups were relatively small it is possible that some of the differences were due to individual variation, such as family traits or the degree of the intellectual impairment of the children. Even if the effect of the chance variation cannot be totally excluded, this would not explain the differences in Portage profiles observed between the groups. In the areas of self-help and motor development, the groups did not differ significantly (except in motor development at age three). Equal development in two areas where the intervention was not supposed to have effect suggests that the basic level of the two groups was not very different. Because the difference was largest in the areas of language and cognitive development, and clear with regard to the social area of the Portage profiles, it seems a warranted conclusion that the signing programme did have effect in these areas.
   Unfortunately, the groups cannot be compared at the earliest ages and for this reason it is not known whether the initial levels of the groups were the same. Comparisons of the intervention group and the research group show that the research group was slightly behind the total intervention group on most measures. It is thus not a 'too good' sample of the total group. In the light of the limited information available, the control group seemed to be slightly ahead of the research group at the age of one year when the first measures were made. Of the five children of the control group who were evaluated at this age, three were among the most advanced in the measures made at the age of five, suggesting that the superiority of the control group may have been due to change variation. At two years of age, seven control children were evaluated. Among them were those two children who later scored lowest on all measures. Thus the average of the control group at two years may have been somewhat too low. However, at later evaluations, the research group was consistently ahead of the control group and the difference remains even if the two low-functioning children of the control group are excluded. When the within-group variation of the two groups is compared, it shows that the research group was ahead as a group. Two or three of the most advanced children in the research group were ahead of the most advanced child in the control group for all Portage areas except motor development at the age of five. Even at five years, in the language area, half of the children in the research group were ahead of the most advanced child in the control group. Lastly, it should be noted that even if observations had been available for all ages, lack of significant differences between the group would not have ensured equal developmental potential since the predictive value of most tests of intellectual functioning infancy is limited (cf. Colombo, 1993).
   The differences between the two groups were the most notable at the age of three years, when the intervention programme was completed. In the course of the two-year follow-up, the differences had started to decrease, but at the age of five years, there were still significant differences between the groups in the areas of language with the help of signs and cognitive development. During the follow-up period, most children received speech therapy once or twice a week, according to individual needs. For the children in the control group, this meant intensified intervention. For the children in the research group it may have meant, to some extent, less intensive intervention. However, it is probable that the early intervention had long-term effects on the interaction style of the whole family. The beneficial communicative environment did, most likely, remain once it was created.
   For most of the children in the intervention group, signing was part of the continuous intervention, at least to some extent. This was also the case for some of the children in the control group. Consequently, the number of manual signs in the control group increased between the ages of three and four years. This increase came mainly from two children for whom signing was emphasized in the individual intervention and whose families had also started to use signed communication. This seems to corroborate the achievements of the research group but also shows that signing may be beneficial even when initiated at an older age. However, according to the results of this study, the vocabulary and the general language development are better when signed communication is incorporated into the child's communication from very early on.

Conclusion
The presuppositions of this study were that children with Down syndrome have specific problems in language acquisition, and particular in developing speech, and that the reasons for these problems lie in difficulties of early interaction and deficient auditory processing. The results of the study show that early intervention with manual signs and special attention to the active communicative role of the child had immediate and long-lasting effects. Children with Down syndrome who participated in the programme used a wider range of communicative means and were more advanced in their linguistic and overall development than the children in the control group. The results support researchers who consider early language intervention to be necessary and important for later achievements and adjustment of children with Down syndrome. They suggest that intervention methods based on manual signing as well as pointing and other pre-speech means of communication may be used similarly to prevent language problems with other groups of children who are at risk for delayed or deviant communication and language development. In this process of development, the people closest to the child play a central role. Manual signing provided the children with a means for active communication and the parents helped the child to create situations in which it could communicate actively and interactively from very early on.

References
     Abrahamsen, A., Lamb, M., Brown-Williams, J. and McCarthy, S. (1991). Boundary conditions on language emergence: Contributions from atypical learners and input. In P. Siple and S.D. Fischer (Eds.), Theoretical issues in sign language research. Volume 2: Psychology. Chicago: Chicago University Press, pp. 231-254.
     Arvio, M., Hautamäki, J. and Tiilikka, P. (1993). Reliability and validity of the Portage assessment scale for clinical studies of mentally handicapped populations. Child: Care, Health and Development, 19, 89-98.
     Ballard, K.D. (1991). Assessment for early intervention: evaluating child development and learning in context. In D. Mitchell and R.I. Brown (Eds.), Early intervention studies for young children with special needs. London: Chapman and Hall, pp. 127-159.
     Bates, E., Camaioni, L. and Volterra, V. (1975). The acquisition of performatives prior to speech. Merrill-Palmer Quarterly, 21, 205-224.
     Bray, M. and Woolnough, L. (1988). The language skills of children with Down's syndrome aged 12 to 16 years. Child Language Teaching and Therapy, 4, 311-324.
     Cardoso-Martins, C. and Mervis, C.B. (1985). Maternal speech to prelinguistic children with Down syndrome. American Journal of Mental Deficiency, 89, 451-458.
     Cardoso-Martins, C., Mervis, C.B. and Mervis, C.A. (1985). Early vocabulary acquisition by children with Down syndrome. American Journal of Mental Deficiency, 90, 177-184.
     Caselli, M.C. and Volterra, V. (1990). From communication to language in hearing and deaf children. In V. Volterra and C.J. Erting (Eds.), From gesture to language in hearing and deaf children. New York: Springer-Verlag, pp. 263-277.
     Casto, G. (1987). Plasticity and the handicapped child. A review of efficacy research. In J.J. Gallagher and C.T. Ramey (Eds.), The Malleability of Children. Baltimore: Paul H. Brookes, pp. 103-113.
     Colombo, J. (1993). Infant cognition. London: Sage.
     Cunningham, C.C. (1988). Early intervention: Some results from the Manchester Down's syndrome cohort study. Paper based on a talk given to The Fundación Catalana Síndrome de Down. Santander, Spain, March 1988.
     Dykens, E.M., Hodapp, R.M. and Evans, D.W. (1994). Profiles and development of adaptive behavior in children with Down syndrome. American Journal on Mental Retardation, 98, 580-587.
     Fischer, M.A. (1987). Mother-child interaction in preverbal children with Down syndrome. Journal of Speech and Hearing Disorders, 52, 179-190.
     Gibson, D. and Harris, A. (1988). Aggregated early intervention effects for Down's syndrome persons: Patterning and longevity of benefits. Journal of Mental Deficiency Research, 32, 1-17.
     Goldin-Meadow, S. and Morford, M. (1990). Gesture in early child language. In V. Volterra and C.J. Erting (Eds.), From gesture to language in hearing and deaf children. New York: Springer-Verlag, pp. 249-262.
     Hornby, G. (1991). Parent involvement. In D. Mitchell and R.I. Brown (Eds.), Early intervention studies for young children with special needs. London: Chapman and Hall, pp. 206-225.
     Johansson, I. (1987). Tecken - en genväg till tal (Signs - a shortcut to speech). Down syndrom: Språk och tal nr. 7. Umeå University, Department of Phonetics. Publication 28.
     Johansson, I. (1988). Språkutveckling hos handikappade barn (Language development in disabled children). Lund, Sweden: Studentlitteratur.
     Johansson, I. (1990). Contributions of language to cognitive development. Paper at The Fourth Biennial Conference of the International Association for Augmentative and Alternative Communication, Stockholm, August 1990.
     Layton, T.L. and Savino, M.A. (1990). Acquiring a communication system by sign and speech in a child with Down syndrome: A longitudinal investigation. Child Language Teaching and Therapy, 6, 59-76.
     Le Prevost, P. (1983). Using the Makaton vocabulary in early language learning with a Down's baby. Mental Handicap, 11, 28-29.
     Levy-Shiff, R. (1986). Mother-father-child interactions in families with a mentally retarded young child. American Journal of Mental Deficiency, 91, 141-149.
     Marcell, M.M. and Weeks, S.L. (1988). Short-term memory difficulties and Down's syndrome. Journal of Mental Deficiency Research, 32, 153-162.
     Maurer, H. and Sherrod, K. (1987). Context directives given to young children with Down syndrome and nonretarded children: development over two years. American Journal of Mental Deficiency, 91, 579-590.
     Mitchell, D. and Brown, R.I. (Eds.) (1991). Early intervention studies for young children with special needs. London: Chapman and Hall.
     Mundy, P., Kasari, C., Sigman, M. and Ruskin, E. (1995). Nonverbal communication and early language acquisition in children with Down syndrome and in normally developing children. Journal of Speech and Hearing Research, 38, 157-167.
     Pine, J.M. (1992). Maternal style at the early one-word stage: re-evaluating the stereotype of the directive mother. First Language, 12, 169-186.
     Pueschel, S.M., Gallagher, P.L., Zartler, A.S. and Pezzullo, J.C. (1987). Cognitive and learning processes in children with Down syndrome. Research of Developmental Disabilities, 8, 21-37.
     Smith, L. and Hagen, V. (1984). Relationship between the home environment and sensorimotor development of Down syndrome and nonretarded infants. American Journal of Mental Deficiency, 89, 124-132.
     Smith, L. and von Tetzchner, S. (1986). Communicative, sensorimotor, and language skills of young children with Down syndrome. American Journal of Mental Deficiency, 91, 57-66.
     Smith, L., von Tetzchner, S. and Michaelsen, B. (1988). The emergence of language skills in young children with Down syndrome. In L. Nadel (Ed.), The psychobiology of Down syndrome. Cambridge, Massachusetts: MIT Press. pp. 145-165.
     Tiilikka, P. and Hautamäki, J. (1986). Portaat - varhaiskasvatusohjelma. Helsinki: The National Welfare Association for the Mentally Retarded. (The Finnish edition of Portage Guide to Early Education.)
     Varnhagen, C.K., Das, J.P. and Varnhagen, S. (1987). Auditory and visual memory span: cognitive processing by TMR individuals with Down syndrome and other etiologies. American Journal of Mental Deficiency, 91, 398-405.
     von Tetzchner, S. and Smith, L. (1986). Mødres tale til tre år gamle barn med Down syndrom (Mothers' speech to three-years-old children with Down syndrome). In P.E. Mjaavatn and L. Smith (Eds.), Barnespråk. Trondheim: NAVF's Senter for Barneforskning, pp. 43-58.
     Windsor, J. and Fristoe, M. (1989). Key word signing: Listeners' classification of signed and spoken narratives. Journal of Speech and Hearing Disorders, 54, 374-382.