Southern African Linguistics and Applied Language Studies 2002 20: 91–103 Copyright © 2002 NISC Pty Ltd

Printed in South Africa — All rights reserved SOUTHERNA FRICANL INGUISTICS

ANDA PPLIEDL ANGUAGES TUDIES
ISSN 1607–3614

Readinga
n dappliedlinguistics—adeafenin
gsilence?1

Elizabeth J Pretorius
Department of Linguistics, Unisa, PO Box 392, Pretoria 0003, South Africa
e-mail: pretoej@unisa.ac.za

A b stra ct:In light of the fact that 2001 has been declared the Year of the Reader, this article pre-
sents some findings concerning reading levels within the learning context, from primary to tertiary
level, and then briefly examines the relationship between reading ability and academic performance.
It is argued that the ability to construct meaning during reading, to access written information inde-
pendently, to acquire, consolidate and utilise knowledge from print information is a strong determi-
nant of academic success. Skill in reading becomes more demanding as students move up the edu-
cation ladder, while the gap between skilled and unskilled readers widens. Unless the reading prob-
lems of our students are addressed, present failure rates will continue. In the light of these findings,
the state of reading research within the South African context is surveyed. In conclusion, questions
concerning a socially responsible applied linguistics are raised, and the implications for teaching and
research are explored.

I n tro du ction
The Minister of Education, Kader Asmal, pro- ing aspects of reading activity whereby written
claimed 2001 to be the Year of the Reader and symbols are translated into language
a national reading campaign called Masifunde (Carpenter & Just, 1986). Comprehension, on
Sonke — Building a nation of readers — has the other hand, refers to the overall under-
been launched to address literacy problems standing process whereby meaning is con-
and to develop a culture of reading and writing. structed within sentence units, between adja-
Given the nationwide focus on reading this cent sentences, and across larger units of text
year, it seems appropriate to undertake some to the meaning of the text as a whole. The inter-
discipline-related ‘stocktaking’ with regard to action between decoding and comprehending
reading research within the South African con- processes in skilled readers happens simulta-
text. Consequently, three main questions will neously and rapidly. Current reading theories
be addressed in this paper, namely: (i) in gen- assume that during this interaction, the reader
eral, what are the reading levels of our stu- builds up a mental representation of the mean-
dents, (ii) how does reading relate to academic ing of the text, which is modified and elaborat-
performance, and (iii) what is the state of read- ed during the reading process.
ing research in South Africa? The paper con- It is commonly acknowledged that compre-
cludes with a discussion of possible reasons for hension cannot effectively occur unless decod-
the present state of reading research in the ing skills have been mastered (Just &
country and considers some implications that Carpenter, 1987; Perfetti, 1988). When children
follow from the findings presented in the paper. start learning to read, emphasis is placed on
the acquisition of decoding skills, and the short
C om po ne ntp roce ss esinre a ding and simple texts that children use in the early
Before continuing, clarification of two central grades are designed to provide maximum prac-
terms related to reading that will be used in the tice of decoding skills such as letter-sound rela-
discussion is called for. A distinction is com- tionships, word recognition and lexical access,
monly made in reading research between the and the syntactic parsing of simple sentences.
two main components of reading, namely Decoding skill is usually reflected in relatively
decoding and comprehension. Decoding fluent oral reading (i.e. reading aloud). Skill in
involves the oculomotor, perceptual and pars- decoding establishes the automatisation of
92 Pretorius: Reading and applied linguistics

lower-level perceptual and linguistic skills guage groups and it is therefore difficult to
which frees the mind, so to speak, for compre- establish accurately what kinds of reading prob-
hension processing. Although decoding is a lems exist and at what level in the educational
necessary reading skill, it is not sufficient, for it process they emerge. Although anecdotal evi-
is comprehension that is the sine qua non of dence abounds on the poor reading levels of
reading. Many readers may decode texts quite primary, secondary and tertiary level students,
readily but still have difficulty understanding this needs to be corroborated with research evi-
what it is that they have decoded (Daneman, dence. Some local research findings concerning
1991; Yuill & Oakhill, 1991). Whether we read the reading levels of our students are discussed
for entertainment, relaxation, study or work pur- in the following paragraphs.
poses, we are not going to be entertained, Although not plentiful, there are several
relax, study nor do our work properly if we do studies that report specifically on reading ability
not understand what it is we are reading. at different levels. For example, the poor read-
Through decoding we ‘learn to read’, while ing skills of black primary school pupils with
comprehension enables us to ‘read to learn’. English as second language (henceforth ESL,
The terms good and poor readers are hence- for convenience — even though English may be
forth used in the sense of good and poor com- the third or even fourth language) are amply
prehenders, irrespective of their decoding documented in the Threshold Project Reports
skills. These terms also exclude from consider- (Macdonald 1990a; b). Although many of the
ation readers who have specific decoding prob- pupils who were tested in this comprehensive
lems, such as students suffering from dyslexia. study were found to be fairly competent at
decoding the English texts they read, their com-
R e ad in gle ve ls inth ele a rn ingc on - prehension levels in general were low, as were
te x t their vocabulary levels and English proficiency
In the domain of reading tests, a distinction is levels. Similarly, in her doctoral research on the
commonly made between three levels of read- reading levels of students for whom English is
ing ability, appropriate to a specific maturational the language of tuition and learning, Strauss
reading level (e.g. Lesiak & Bradley-Johnson, (1995) reports that the black ESL Grade 6 stu-
1983): dents that she tested had good decoding skills
* At the independent level, the reader should in English but very poor comprehension skills
read with 98% decoding accuracy and have (less than 30% comprehension levels). Both
at least a 95% level of comprehension. Macdonald (1990a) and Strauss (1995) report
These are highly skilled readers who, as the that there is a tendency for children in histori-
name indicates, can independently access cally disadvantaged (i.e. black) schools to “bark
information from texts and can effectively at print”, i.e. to become what Devine (1988)
learn from texts appropriate for that specific calls “sound-centred readers”, where the peda-
maturational level. gogic focus is on getting readers to decode
* At the instructional level the reader should printed information with little attention paid to
read with 95% decoding accuracy and 75% the meaning of the passage. More recently, the
comprehension. These are readers who do headlines of the Sunday Times (16 July 2000)
not have major reading problems but who proclaimed South African children to be the
can benefit from reading instruction at their “dunces of Africa”. The feature article reported
maturational level. on the findings of a comparative study of litera-
* At the frustration level, the reader reads with cy and numeracy rates of primary school chil-
less than 90% decoding accuracy and 60% or dren from 12 countries in Africa, with South
less comprehension. These are readers who African children in general faring poorly in com-
have major reading problems, especially with parison to their African peers on both literacy
regard to comprehending written information, and numeracy measures.
and who are reading well below their matura- Such results are not confined to the prima-
tional level. They need intensive reading pro- ry school. Although research on reading levels
grammes to increase their reading level. at secondary school are sparser, the findings
In South Africa there are no standardised are equally dismal. The READ Annual Report
reading tests for different maturational or lan- (1999) reports that in rural areas, while the
Southern African Linguistics and Applied Language Studies 2002. 20: 91–103 93

average age of entry of Grade 8 pupils is 14.4 below the recommended minimum speed of at
years, on average they have ESL reading lev- least about 160 words per minute for ESL read-
els equivalent to children at age 7.6 years. ers. Reading research suggests that reading at
Every year there is a public outcry over the low too slow a rate not only reduces enjoyment of
matriculation marks and poor pass rates coun- reading but also jeopardises efficient compre-
trywide of our secondary school students, par- hension (Anderson, 1999). This was clearly evi-
ticularly students who have to study through the dent amongst these readers for not only did
medium of a language which is not their prima- they read slowly, they also averaged a 40%
ry language. Although the reading levels of comprehension rate.
these matriculants are not stated, their high fail- There are of course numerous historical,
ure rate suggests problems in ‘reading to learn’. socio-political and cultural factors that have
Poor reading levels are evident at tertiary contributed to these nationally low reading lev-
level too. For example, the 1989 results of els. The deleterious policies and practices of
reading tests at the University of Natal, apartheid education gave rise to an extremely
Pietermaritzburg, indicated that Black ESL stu- dysfunctional education system characterised
dents had average reading rates of 174 words by poor teaching, reliance on rote teaching and
per minute (wpm)and a 62% level of compre- learning, low proficiency levels in the medium
hension, while White L1 students had average of tuition, overcrowded classrooms, poor
reading rates of 246 wpm and a 76% level of school management, low levels of professional-
comprehension (Blacquiere, 1989). Souter, ism and few resources. Needless to say, these
Archer and Rochford (1992) found that stu- factors were not conducive to the development
dents at a teacher’s training college had gener- of sound reading skills and practices within the
ally low levels of reading and fared much better schooling system. Furthermore, within our edu-
on literal than inferential questions. The picture cational context, during the ‘learning to read’
is even bleaker at the historically disadvan- stage from Grades 1–3, there was, and still is,
taged universities. Perkins reports that the an overemphasis on decoding skills, with very
1989 intake of students at the Student little attention given to the development of com-
Orientation Programme at the University of prehension skills, and there was and still is a
Transkei were given the Stanford Diagnostic gross shortage of a variety of age-related suit-
Reading Test: “Results indicated that only able reading texts and textbooks on which
13,8% of the students had the reading skills learners can practise and hone their reading
necessary to comprehend textbooks for first- skills in both their primary and additional lan-
year students. Twenty-six per cent were found guage. Compounding this situation is the fact
to be unable to cope without assistance” that many students come from a largely oral
(Perkins, 1991: 232). At the Naete conference culture in which reading plays a very small role,
in 1999, Webb (1999) reported that many of the children are not exposed to storybook reading
ESL first-year students at the University of or other literacy practices from an early age,
Pretoria who were tested had reading levels of and reading is not perceived to be a meaning-
Grade 7–8 students. Using inferential ques- ful leisure activity. There is also very little
tions as an index of reading ability, Pretorius research on the reading problems of Black stu-
(2000) found that the average reading level of dents in their primary languages, due no doubt
first year Psychology and Sociology students at to the fact that little reading is done in the
Unisa was around 53%, which means that the African languages within the schooling context.
majority of students were reading at frustration Furthermore, many of these students also
level, i.e. well below their assumed reading come from socio-economically disadvantaged
level. The first-year Medical and Occupational homes, a variable which worldwide is typically
Therapy students at Medunsa for whom associated with low reading levels and acade-
English is the medium of learning and tuition mic underperformance. It is no wonder that so
fared marginally better at a mean 57% compre- many of our students have poor reading skills -
hension level (Pretorius, 2000). The average in fact, it is surprising that many of them cope
reading speed of a small sample of first-year as well as they do, given the factors that mili-
Unisa ESL students was also tested (N=24), tate against the proper development of their
and this was a mere 96 words per minute, far reading skills in particular.
94 Pretorius: Reading and applied linguistics

From the studies cited above, it is evident and academic performance in our own complex
that reading is a major problem in the South local situation?
African learning context, especially for ESL stu- Pretorius (1998) tested and observed 24
dents for whom English is the language of Grade 4 pupils at a private primary school for
learning and tuition. But what exactly is the boys over a period of six months, and com-
relationship between reading ability and acade- pared their reading ability on both narrative and
mic performance? Do low reading levels nec- expository texts with their academic perfor-
essarily comprise a barrier to effective learn- mance. Although there were no formal mea-
ing? These are issues that are addressed in the sures of the pupils’ decoding skills, impression-
following section. istic judgements were made for each child dur-
ing the reading of narrative texts, based on mis-
R ea d inga n da cade m icp erfor m a n ce cues (sounding wrong word), ease of word
Because the majority of students in South recognition, ability to sound out unfamiliar
Africa study through the medium of a language words, and phrasing patterns (i.e. whether the
which is not their primary language, it is com- child read word-by-word or used larger phras-
monly assumed and argued that the low levels ing units). Comprehension was measured by
of ESL proficiency of our students play a signif- means of inferential questions on the text and
icant role in low academic performance. While ability to resolve anaphoric references. The
not disputing the relationship between lan- reading scores revealed four distinct groups of
guage proficiency and academic performance, reading ability levels within the class (Table 1).
in this paper it is argued that a fundamental fea- The boys in Group I performed well below
ture of academic underperformance in South the rest of the class. They performed poorly on
Africa is poor reading ability, but that this factor all the different reading measures, both decod-
tends to be overshadowed by the language ing and comprehension, in both the narrative
proficiency issue and other educational woes. and expository genres. The oral fluency of the
Overseas research findings in applied lin- pupils in Group II tended to be slow and halting
guistics and reading research consistently with several miscues, and their decoding skills
show a strong correlation between reading pro- had clearly not yet been automatised. Although
ficiency and scholastic or academic success at this group of boys performed well in making
all ages, from the primary school through to ter- inferences and resolving anaphors in the narra-
tiary level: students who read a lot and who tive text, their performance dropped by as
understand what they read are usually students much as 20–30% when it came to making infer-
who obtain good grades (e.g. Heath, 1983; ences in the expository texts. Their compre-
Snow, 1983; Wells, 1986; Elley, 1991; Krashen, hension skills were not yet robust enough to
1993). These results obtain both for students enable them to accurately and effectively make
who read and study in their primary language sense of the more unfamiliar genre of exposito-
and for those who read and study in an addi- ry texts — the kind of texts they would increas-
tional language. In fact, many researchers now ingly need to read to learn from. The pupils in
argue that reading also significantly improves Group III comprised a middle or intermediate
language proficiency, and that reading skills are group of readers who did not seem to have any
the most important skills that additional lan- major problems when reading and understand-
guage learners need in the learning context ing texts. The differences between their perfor-
(Saville-Troike, 1984; Carrell, 1989; Hafiz & mance in the narrative and expository texts was
Tudor, 1989; Elley, 1991; Mbise, 1993). But narrower than Group II, suggesting better inde-
what about the relationship between reading pendent accessing of information from texts

Table 1: Academic performance and mean reading scores of Grade 4 students

Reading ability group I II III IV

Mean reading score % 49% 63,8% 75,4% 88,5%
N 2 8 9 5
Academic Performance low achievers average achievers good achievers top achievers
Southern African Linguistics and Applied Language Studies 2002. 20: 91–103 95

than the previous group. Despite occasional dependent variable. The reading scores, which
hesitancy in terms of oral fluency, on the whole served as the independent variable, were then
this group of subjects displayed fairly good compared to the academic groupings, as shown
decoding and comprehension skills in both the in Table 2. The reading results showed a very
narrative and expository genres. The pupils in clear relationship between academic perfor-
Group IV consistently performed well in all the mance and reading ability. Here too, as can be
tests and on all the different reading measures, seen, clear differences emerged in reading abil-
and their scores were consistently higher than ity across the different academic groups.
the other subjects. They were clearly skilled A multiple regression analysis was then
readers and could independently access infor- applied to see which sets of data (i.e. reading
mation from either narrative or expository texts. ability, matric results, study habits, level of moti-
Their oral reading was also notably fluent and vation, locus of control) best predicted acade-
expressive. mic performance in the final psychology exam-
At the end of the year, after the reading ination (Kruger & Janeke, 2000). The results
tests had already been conducted, the class showed that, of all the variables, the reading
teacher was asked to group the class into four score was the strongest predictor of academic
groups according to overall academic perfor- performance (r = 0.48, p < 0.05, R2 = 0.24),
mance, viz. low, average, good and top achiev- accounting for 24% of the variance. The sec-
ers. The reading scores predicted 22 of the 24 ond strongest predictor was locus of control,
children’s academic performance (there were which accounted for 15% of the variance, while
two children who, academically, were in the matric results only accounted for 0.7% of the
‘good achievers’ group but whose reading variance (Kruger & Janeke, 2000).
scores were in Group II). In other words, the What these sets of findings from both pri-
findings showed a strong relationship between mary and tertiary level indicate is that there is a
reading ability and academic performance — strong relationship between reading ability and
the level at which the children were reading academic performance. Students who score
predicted the level at which they were perform- poorly on reading tests are generally students
ing academically. who perform poorly academically. The better
Moving on to tertiary level, we see a similar the reading ability of students, the better their
pattern. Using inference test items as an index chances of coping academically. This relation-
of reading ability, Pretorius (2000) tested 1 200 ship starts early in the primary school and con-
undergraduate Psychology at Unisa. The read- tinues up to tertiary level, and applies equally to
ing test, compiled by Pretorius, was part of a students who study in their primary or in an
larger battery of tests compiled by the additional language.
Department of Psychology, which included These results are not surprising. Reading
items relating to factors such as study habits, promotes the development of meaning making
level of motivation, locus of control and matric and information processing abilities that are
results. On the basis of the marks the students valued in the learning context and that are
received in their final Psychology examination, indeed necessary for success in the technolog-
they were divided into five academic groups, ical and information age in which we live.
viz. Fail I (0–39%), Fail II (40–49%), At Risk Constant and regular exposure to print enables
(50–59%), Pass (60–73%) and Distinction (73% reading processes to become automatised and
or more). The academic groups served as the hence more effective, and also provides the

Table 2: Academic performance and mean reading scores of undergraduate psychology students a

Academic groups Fail 1 Fail II At Risk Pass Distinction

(0–39%) (40–49%) (50–59%) (60–72%) (73+%)
N 355 283 185 178 132
Mean reading score 44.6 50.4 55.2 60 70.1

a The mean scores are expressed as percentages

96 Pretorius: Reading and applied linguistics

reader with opportunities for building up new that they do manage. In spite of their persever-
knowledge structures derived from the informa- ance in their studies, many students continue to
tion in the text. Differences in reading ability perform poorly and never really ‘grow rich’, lin-
create what has been called Matthew effects2 guistically, cognitively, textually, or in terms of
(Stanovich, 1986), whereby ‘the rich get richer subject-related knowledge, one of the major
and the poor get poorer’. Matthew effects dis- reasons being that they are poor readers and
play the following negative cyclic patterns: stu- hence poor processors of information, and this
dents with difficulties in reading read less than inhibits academic success because they have
good readers. Because they read less, the pro- problems reading to learn.
cessing mechanisms are exercised less, and It is time to openly acknowledge that many
subsequently the automatised cognitive-lin- South African students at all levels are proving
guistic process and skills that support compre- to be dismally unskilled readers and that this
hension are not as well developed as good has serious consequences for their academic
readers. Because reading is a frustrating and performance. Language proficiency and reading
effortful task, weak readers lose the motivation ability are undeniably related, especially in an
for reading, which in turn affects the amount of L2, but they are discrete skills despite their com-
reading practice they get. Their teachers and plementarity. The development of ability in spo-
parents also tend to have lower expectations ken language does not automatically lead to the
for these pupils. Lowered expectations, low- development of ability in written language. If we
ered motivation, lowered self-esteem and low- wish to turn the educational crisis in our country
ered levels of practice and exposure to print around, then we need to address the reading cri-
aggravate reading problems. And so the nega- sis within our educational system and examine
tive cycle continues. As Spear-Sterling and more closely the extent, the nature, the causes
Sternberg (1996: 9) point out, “... children who and the consequences of our students’ reading
start off poorly in reading rapidly become even problems. Since language is the interface
more disadvantaged relative to other readers, between thinking and the world, and because
whereas the reverse happens for children who we construct meaning largely through language
have a successful start in reading”. to make sense of our world, it is important for us,
Matthew effects in reading have a spill-over as linguists and applied linguists, to give greater
effect in the learning context. In other words, recognition to the fact that meaning construction
students who exhibit Matthew effects in their (and hence also learning) via the oral and written
reading also tend to exhibit Matthew effects in modes are related yet distinct skills. Through
their academic performance in general. When reading research linguists and applied linguists
one considers the broader South African socio- can gain a better understanding of the linguistic-
political and educational context through which cognitive nature of academic underperformance
many of our students have moved, it is clear in general, and in the South African context in
that they start off with poorly developed reading particular, in order to contribute to the national
skills and they continue to read at a suboptimal debate and make well-informed decisions with
level throughout their schooling career, never regard to normalisation, remediation and pre-
catching up. They seldom associate reading ventative measures.
with reading for pleasure or relaxation, and they
generally experience reading as an effortful T h es ta te o f re a d ing re se a rc h in
task, thus developing negative attitudes towards S o u thA frica
it. As they progress through the schooling sys- Judging from the sheer volume of literature and
tem, they experience problems understanding research published in reading in journals world-
the language and conventions of the textbooks wide, reading is a research domain in its own
they read, and because they experience read- right but it also straddles several disciplines, the
ing as effortful, they have little motivation to most obvious being linguistics, psychology and
read, so they do little reading beyond what they education. If, as (Tuckman, 1988: 14) states,
are required to do. As a result, their language “research has value in contributing to knowl-
never develops beyond their oral competence, edge”, then, by looking at the amount of
they have difficulty reading to learn and they research being done in a particular domain one
never seem to gain much from the little reading can assess the extent to which it is contributing
Southern African Linguistics and Applied Language Studies 2002. 20: 91–103 97

to knowledge in that particular area. Given the reading in some way or another. Of the remain-
importance of reading in the learning context, ing 235 articles, five articles refer to reading in
and given too that students’ reading levels are a subheading or in passing, in five articles the
generally poor in SA, one would expect there to importance of reading is implied but not explic-
be a fair amount of research in this area in itly discussed, while in five articles that look at
applied language studies in South Africa. factors affecting underpreparedness or low
Consequently, a small scale literature survey ESL proficiency levels, reading as a variable in
was undertaken of the amount of articles on academic performance is not mentioned at all.
reading published in South Africa from 1990 to In all the journals, especially in the SAJHE,
the present day. Four journals that were thought but notably with the exception of the SAJL,
to be likely to publish matters relating to applied there have been numerous articles that
linguistic issues were selected for the survey, describe, discuss and debate issues relating to
viz. the South African Journal of Applied academic literacy and academic development
Language Studies (SAJALS) which appeared in general (e.g. Boughey, 1998; Amos, 1999;
twice a year,3 the Journal for Language Von Gruenewaldt, 1999). In the South African
Teaching (SAALT), a quarterly journal, the context, ‘academic development’ is a two-way
South African Journal of Higher Education business, involving capacity building amongst
(SAJHE) which appears three times a year, and both students and tertiary level lecturers. In the
the South African Journal of Linguistics (SAJL), journal there are several articles that assess
a quarterly journal. Although the SAJHE is not a the skills and abilities of undergraduate stu-
linguistic journal, it was included because it is dents (e.g. Miller, 1997; Kilfoil, 1999) or the effi-
an important journal forum that specifically cacy of specific academic development pro-
addresses issues relevant to tertiary level learn- grammes that have been implemented at vari-
ing and teaching in South Africa, and which pub- ous universities or technikons around the coun-
lishes an average of 25 articles per edition. try (e.g. Miller et al., 1997; De Villiers &
Because this was a ‘reconnaissance’ and not a Rwigema, 1998). There are also several arti-
comprehensive survey, journals directed at pri- cles outlining how English courses have
mary and secondary school issues were not changed in response to the needs of ESL stu-
included, nor was the supplement to the LSSA dents (e.g. Wood, 1997), shifting from a strong
journal included, which often addresses applied literary emphasis to a language emphasis in
linguistic issues. The figures relating to articles which reading, writing, text and discourse skills
published on reading during the past decade in appropriate to the academic context are devel-
the four journals are shown in Table 3. oped or critical language awareness promoted.
As can be seen from this table, out of a total There are some articles that deal specifically
of about 1 202 articles, only 17 pertinently deal with the teaching of writing (e.g. Moyo, 1995;
with reading issues, comprising 1.4% of the Orr, 1995) or the importance of teaching Critical
article output. From this it is clear that not much Language Awareness or literacy as critical
research is being done in SA in the field of social practice (e.g. Boughey, 1998; Von
reading, at least not in journals that publish arti- Gruenewaldt, 1999). But in none of these arti-
cles related to linguistics and applied language cles is there even a heading or a subheading
studies. For example, of the 245 articles in the dealing with reading. Many of these articles
SAALT journal, only 10 pertained directly to certainly refer to the fact that students lack

Table 3: Amount of articles on reading in four South Africa journals: 1990–1999

SAJALS SAALT SAJHE SAJL

(x2) (X4) (X3) (X4) Totals
Total articles (1990–1999) +57 245 +700 200 1 202
Articles on reading (1990–1999) 5 (8.7%) 10 (4%) 2 (0.2%) 0 17

SAJALS = South African Journal for Applied Language Studies (1992–1999) 3; SAALT = Journal for Language
Teaching; SAJHE = South African Journal for Higher Education; SAJL = South African Journal of Linguistics
98 Pretorius: Reading and applied linguistics

reading skills, and many refer to the importance academic development programme — it consti-
of reading at tertiary level, but they do so tutes the very process whereby learning occurs.
almost in passing. Miller (1997: 17) comes clos- I would say the dearth of research in this area
est to hitting the nail on the head when he amounts to a rather deafening silence.
states, in his final paragraph, that the results of
his study suggest that “before we embark on D is cu ss io n
programmes that teach students how to write, In the South African learning context, ESL poor
we need to ensure that they know how to read”. academic performers are also usually poor
Judging from the articles in these journals, readers, and if we wish to help these students
the debate on the nature of academic literacy then we need to recognise at a very profound
and the various directions that academic devel- level that such students have reading problems
opment programmes should take is certainly that impede the development of their academic
alive and well. The numerous external factors potential. Reading problems do not go away if
that contribute to and cause academic under- left unattended. In fact, the gap between skilled
preparedness are readily acknowledged, such and unskilled readers usually widens as stu-
as historical, socio-political, cultural and eco- dents move up the educational ladder (Chall et
nomic factors, the low levels of literacy and al., 1990). Unless the reading levels of our stu-
high levels of poverty, the poor schooling during dents are improved, they will not develop the
apartheid and the highly politicised and disrupt- requisite linguistic-cognitive skills for construct-
ed schooling during the 1980s until political lib- ing meaning so that they can read to learn,
eration in 1994. Several of the articles also deal Matthew effects will continue to dog their
with the numerous variables that typically make attempts at becoming academically literate,
up the profile of an underachieving student, and they will continue to have limited and inef-
such as personal circumstances, study meth- fective access to the rich sources of information
ods and habits, time management, use of the from which they are supposed learn.
library and computers, motivation, interest, past Reading research is only starting to uncov-
schooling experiences, performance in the final er the powerful feedback effects from print
matriculation examination, and ESL proficiency. exposure on the development of cognitive-lin-
Although more communicative and text- guistic abilities that underpin language devel-
based approaches to ESL pedagogy are to be opment and language comprehension in partic-
welcomed, they do not necessarily improve ular, and academic performance in general.
reading skills. While writing is integral to the Reading not only improves reading skill, it also
academic context, students cannot become bet- promotes language development, and has
ter writers unless their reading skills improve. been found to be a powerful means for improv-
The printed word and the information contained ing proficiency in an additional language, espe-
therein that are accessed through reading pro- cially the cognitive-academic language profi-
vide the role models for writing. And although ciency that is highly valued in the learning con-
Critical Language Awareness has an important text (cf. Elley, 1991; Feitelson et al., 1993;
role to play in the academic world, expecting Mbise, 1993).
students to develop critical language skills when
they are probably reading at frustration level is R e as on sfo rlim ite dr ea din gre s e arc h
putting the cart before the horse. What is strik- w ithinth eS o uthA fricanc o nte xt
ing from a cursory survey of the articles in this One can speculate about the many reasons
journal during the past decade is the general, why reading is receiving so little attention in
almost perfunctory, vague acceptance that South Africa. One reason may lie in our failure
reading is important, but an absence of close to acknowledge the fundamental role that read-
empirical investigation and, concomitantly, ing plays in the mediation of knowledge and
meaningful debate and discussion at a theoreti- learning. The neglect of reading in the class-
cal, methodological and pedagogical level as to room may stem from the fact that the four lan-
the nature of reading problems and considera- guage skills are commonly assumed to be
tion of what is being, and should be, done about equally important, viz. listening, speaking,
them. To reiterate, reading is not simply an addi- reading and writing. The objective of both pri-
tional tool that students need to master in an mary and secondary language courses is to
Southern African Linguistics and Applied Language Studies 2002. 20: 91–103 99

promote proficiency in all four skill domains, erences of 238 TESL students at four colleges
such that reading is relegated to just another and university, Mawasha et al. (1994a; 1994b)
language skill the learner needs. However, in found that only 5% of the students did any plea-
the learning context, I would suggest that there sure reading, while only 4.6% saw reading as
is a hierarchical ordering to these skills and that improving on medium of instruction in the TESL
from about Grade 4 onwards, proficiency in classroom.
reading becomes the most important skill that A third reason for overlooking reading lies in
determines both further language development its ubiquitous nature — because it is one of the
and academic success. Because reading com- basic skills that is taught in early primary
prehension depends largely on a reader’s pre- school, once it is taught, skill is assumed, and
cision and skill in attending to and utilising lin- because it is a taken-for-granted skill, it is eas-
guistic clues in the text for constructing mean- ily overlooked. Paradoxically, everyone says it
ing, reading promotes the ability to use and is important but beyond Grade 2 or 3, little is
understand language in an increasingly disem- actually being done about it in the classroom.
bedded context, viz. that of the printed word. In Besides ploughing through the prescribed
other words, reading promotes the develop- books during English periods, there are seldom
ment of cognitive-academic language proficien- other periods assigned to the teaching of read-
cy (CALP). The cognitive-linguistic skills that ing skills (let alone the teaching of reading
develop as a result of exposure to written lan- across the curriculum) — and if such periods
guage are the kinds of skills that are crucial in are set aside for reading, then they are often
the learning context because they enable learn- used for doing homework, or nothing, or for
ers to access, understand and learn from print catching up on other subjects. Reading may
information autonomously, rapidly and effec- also be equated with decoding, merely a tech-
tively, and in an increasingly abstract way. Yet nical skill, and not with the more fundamental
there are many students, particularly those who function of meaning making. On the mistaken
have to study through the medium of an ESL, assumption that if students can technically read
who struggle through their entire schooling (i.e. decode), then they don’t have a reading
career, one of the major reasons being that problem, little further attention is given to devel-
their reading skills are poor, hence they cannot oping higher order reading skills in students,
effectively read to learn and they cannot and the reading comprehension problems go
become independent learners. Yet this need undetected and ignored. In addition, the fact
not be the case, for reading skills can be that we have no standardised reading tests for
improved through explicit reading instruction different maturational levels means that we
and regular exposure to a variety of reading cannot assess national reading levels and
materials. The implicit assumption that lan- detect problems. This makes it easier for us to
guage proficiency is the gateway to reading be in denial about the reading problem within
ability deflects attention away from reading the educational system. Furthermore, when
problems and the need for explicitly developing students have difficulty reading to learn, it is
and nurturing reading ability. usually assumed that their comprehension
Another possible reason why reading has problems stem from limited language proficien-
received so little attention is due to the low cy. Yet, as research findings in the past decade
value attached to reading by our learners and have shown, we need to reconsider the sup-
by the wider, largely oral, community from posed directionality of this causal relationship:
which they come — reading is often equated increased language proficiency does not nec-
with rote learning, it is not perceived as a essarily lead to increased reading comprehen-
meaning-making tool to acquire new knowl- sion ability, while increased reading compre-
edge, and it is seldom equated with pleasure. hension ability does lead to increased language
Yet it is through pleasure reading in the proficiency (e.g. Daneman, 1991; Elley, 1991;
younger grades that students refine their read- Hacquebord, 1994).
ing skills, which in turn enable them to compre-
hend the more complex expository texts from To wa rdsas ociallyr e spon siblelinguistics
which they learn in their later schooling years. Students with reading problems get caught in a
For example, in their study of the reading pref- negative cycle of failed reading outcomes and
100 Pretorius: Reading and applied linguistics

non-strategic reading styles, and if they are in academic underpreparedness and establish
reading at frustration level and have to deal the cardinal status of reading research. We
with texts well above their reading ability level, need to play an active role in fostering a culture
then leaving them to their own devices in the of reading within our multilingual and multicul-
hope that problems will sort themselves out tural context, in making students and educators
amounts to an abdication of educational aware of the important role reading plays and of
responsibility. If the socio-economic and cultur- the severe academic consequences that poor
al context of the students fails to support the reading ability carries. Some methodological
acquisition of literacy practices that result in and pedagogical implications that follow from
skilled reading, and if the primary and sec- the above discussion are:
ondary school contexts have failed to ensure • Reading is a valid and relevant domain for
the development of these skills, then tertiary language and linguistic studies. The complex
institutions are hard pressed to justify turning a relationship between reading and other vari-
blind eye to the reading problems of their stu- ables such as academic performance, lan-
dents, especially distance education institutions guage proficiency, vocabulary levels, and
that rely par excellence on print-based material reading comprehension and listening com-
as a means of learning and acquiring new skills prehension need to be further explored. Other
and knowledge. With the ever-increasing trend issues that require further research include an
towards distance learning institutions that is examination of the relationship and transfer
emerging all over the country, the need for effects between reading in a primary and
skilled reading is particularly compelling, given reading in an additional language, the effects
that most of the learning occurs through the of pleasure reading on reading ability, the
medium of the printed word rather than the spo- establishment of family literacy programmes
ken word. within schools and communities, the effects of
Typical responses at tertiary level to the exposure to books from an early age and sub-
poor academic performance of these students sequent reading skill and academic perfor-
have included the reduction and simplification mance, and so on. At a more linguistic level,
of course contents, the introduction of academ- the role of syntactic parsing, vocabulary and
ic development programmes to promote lan- conceptual development, anaphoric resolu-
guage and study skills for academically vulner- tion, and inferencing during reading are all
able students, and a strong onus on lecturers to issues that require further analysis.
write or select less complex, more reader- • Modules/courses on reading should be intro-
friendly texts for their students. While all these duced into our academic programmes in lan-
measures are not without their merit, it is impor- guage, linguistic and education departments
tant not to simply treat the symptoms of poor to make students aware of the component
reading ability, but also to treat the cognitive-lin- processes involved in decoding and text
guistic causes. Reading is not an additional tool comprehension, the developmental stages in
that students need to master in the learning learning to read and write and the nature of
context — it constitutes the very process the relationship between language, literacy
whereby learning occurs. and academic performance.
Having a better understanding of the nature • Reading was badly neglected in the previous
of the reading problems experienced by stu- government’s educational policies and prac-
dents permits us greater insight into the cogni- tices and we need to ensure that history does
tive-linguistic dynamics of meaning construc- not repeat itself. We therefore need to critical-
tion and the dynamics of academic perfor- ly examine the role of reading in outcomes-
mance, especially the continued poor academ- based education, in theory and practice.
ic performance of many students, which in turn • The training of ESL college students in
enables us to make more informed decisions understanding the reading process and in
about the short- and long-term solutions we improving their comprehension skills needs
seek to address these problems. to assume top priority. Teachers with low
If applied linguistics is to be a socially reading levels may teach students to read
responsible discipline, then as applied linguists superficially and so perpetuate the negative
we need to recognise the importance of reading cycle of Matthew effects in our schools.
Southern African Linguistics and Applied Language Studies 2002. 20: 91–103 101

• The promotion of reading for pleasure within should involve practical, hands-on ways of
the educational context should assume prior- developing reading skills, and make use of
ity, since it is through pleasure reading and authentic expository texts that relate to the
constant and regular exposure to a variety of subjects that the students are studying.
written texts that skill in ‘reading to learn’ is To conclude, the Year of the Reader comes
developed. at a timeous moment in our educational history.
• Principals, teachers, parents, those in early Whether institutions, educationalists and
child development and other educational researchers will heed the call for greater atten-
stakeholders should be made aware of the tion to be paid to reading remains to be seen.
important role of reading and books in pro- Because language is the medium through
moting and mediating learning, language which reading is mediated, departments of lin-
proficiency and academic success. guistics, applied linguistics or applied language
• Despite financial constraints, primary, sec- studies are departments that one would expect
ondary and tertiary level institutions should to be at the forefront of reading research in the
provide services that can help students test rich and complex multilingual and multicultural
their reading abilities and develop their read- South African context.
ing skills. Reading awareness and the devel-
opment of reading skills should be a core N o te s
component of foundation courses, of acade- 1 The article is based on a paper of the same

mic development programmes and of English title read by the author at the Southern
ESL courses. Ancillary reading ‘clinics’ African Applied Linguistics Association
should also be available to help students on (SAALA) Conference, held at the University
an individual level. Consideration should also of Stellenbosch, 14–16 September 2000. The
be given to designing diagnostic reading theme of the conference was: Towards a
tests suitable for assessing the reading levels socially responsible applied linguistics.
of students in the learning context. These 2 The term comes from the book of Matthew in

tests should not be seen as gate-keeping the New Testament, 25:29: For unto every-
devices but as tools for making students one that hath shall be given, and he shall
aware of their reading level and of what have in abundance, but from him that hath
needs to be done to develop their reading not shall be taken away even that which he
skills. Reading awareness and skill develop- hath.
ment should also be built into individual con- 3 This journal (SAJALS) was founded in 1992.

tent courses. Building capacity in this way It was discontinued in 2000 and has now
can prevent academic failure in the long term amalgamated with the South African Journal
and so justify expenditure in this area. of Linguistics as Southern African Linguistics
Whatever form this capacity building takes, it and Applied Language Studies.

Re fere n c es
Amos TL. 1999. Integrating the development courses for academic development practi-
of academic literacy into mainstream teach- tioners. South African Journal for Higher
ing and learning. South African Journal Education 12: 166–171.
for Higher Education 13: 177–183. Carpenter PA & Just MA. 1986. Cognitive
Anderson NJ. 1999. Improving reading speed: processes in reading. In: Orasanu J (ed),
Activities for the classroom. English 1986. Reading Comprehension: From
Teaching Forum 37: 2–5. Research to Practice. Hillsdale: Lawrence
Blacquiere A. 1989. Reading for survival: text Erlbaum.
and the second language student. South Carrell PL. 1989. SLA and classroom instruc-
African Journal of Higher Education 3: tion: Reading. Annual Review of Applied
73–82. Linguistics 9: 233–242.
Boughey C. 1998. Language and “disadvan- Chall JS, Jacobs VA & Baldwin LE. 1990.
tage” in South African institutions of higher The reading crisis: Why poor children fall
education: Implications of critical chal- behind. Cambridge, Mass.: Harvard
lenges to second language acquisition dis- University Press.
102 Pretorius: Reading and applied linguistics

Daneman M. 1991. Individual differences in Learning Experiences. Pretoria: HSRC.

reading skills. In: Barr R et al. (eds), Mawasha A et al. 1994a. Student objectives in
Handbook of Reading Research Volume II. reading. Journal of Language Teaching
London: Longman. 28(1).
De Villiers J & Rwigema H. 1998. The effect of Mawasha A et al. 1994b. Reading preferences
a bridging year on the graduation success amongst TESL students. Journal of
of educationally disadvantaged commerce Language Teaching 28(2).
students. South African Journal for Mbise A. 1993. Teaching English language
Higher Education 12: 103–108. reading in Tanzanian secondary schools. In:
Devine J. 1988. A case study of two readers: Rubagumya CM (ed), Teaching and
models of reading and reading perfor- Researching Language in African
mance. In: Carrell PL, Devine J & Eskey DE Classrooms. Clevedon: Multilingual
(eds), Interactive Approaches to Second Matters, pp 96–109.
Language Reading. USA. Miller R. 1997. Mark my words, part 2:
Elley WB. 1991. Acquiring literacy in a second Students. South African Journal for
language: The effect of book-based pro- Higher Education 11: 11–18.
grammes. Language Learning 41: Miller R, Bradbury J & Wessels SL. 1997.
375–411. Academic performance of first and second
Fasheh MJ. 1995. The reading campaign language students: Kinds of assessment.
experience within Palestinian society: South African Journal for Higher
Innovative strategies for learning and build- Education 11: 70–79.
ing community. Harvard Educational Moyo T. 1995. Student academic writing in an
Review 65: 66–92. ESL situation. South African Journal of
Feitelson D, Goldstein Z, Iraqi J & Share DL. Higher Education 9: 168–172.
1993. Effects of listening to story reading on Orr MH. 1995. Teaching writing at university:
aspects of literacy acquisition in a diglossic an academic literacy programme. South
situation. Reading Research Quarterly 28: African Journal of Higher Education 9:
71–79. 189–197.
Hacquebord H. 1994. L2 reading in the con- Perfetti CA. 1988. Verbal efficiency in reading
tent areas: text comprehension in sec- ability. In: Daneman M, Mackinnon GE &
ondary education in the Netherlands. Waller TG (eds), Reading Research:
Journal of Research in Reading 17: Advances in Theory and Practice Vol 6.
83–98. New York: Academic Press.
Hafiz FM & Tudor I. 1989. Extensive reading Perkins DM. 1991. Improvement of reading
and the development of language skills. and vocabulary skills at the University of
English Language Teaching Journal 43: Transkei. South African Journal of
4–11. Education 11: 231–235.
Heath SB. 1983. Ways with Words. New York: Pretorius EJ. 1998. Reading and inferencing
Cambridge University Press. skills amongst ten-year-olds during the
Janeke HC & Kruger P. 2000. Assessing and reading of narrative and expository texts.
predicting student academic performance: Unpublished report submitted to Waterkloof
The problem of measurement. Reports from House Preparatory School, Pretoria,
the Psychology Department, Unisa. November 1998, (based on the reading and
Just MA & Carpenter PA. 1987. The inferencing abilities of Grade 4 pupils at
Psychology of Reading and Language WHPS, Pretoria, April–November 1998.)
Comprehension. Boston: Allyn & Bacon. Pretorius EJ. 2000. Inference generation in
Kilfoil WR. 1999. The linguistic competence of the reading of expository texts by university
science students. South African Journal students. D Litt et Phil. Thesis, University of
for Higher Education 13: 47–55. South Africa, Pretoria, South Africa.
Krashen S. 1993. The Power of Reading. READ Educational Trust 1999. Annual
Englewood, Colorado: Libraries Unlimited. Report. Braamfontein, South Africa.
Lesiak J & Bradley-Johnson S. 1983. Reading Saville-Troike M. 1984. What really matters in
Assessment for Placement and Programming. second language learning for academic
Springfield: Charles C. Thomas. purposes? TESOL Quarterly 18: 199–219.
Macdonald CA. 1990a. Crossing the Snow CA. 1983. Literacy and language: rela-
Threshold into Standard Three. Pretoria: tionships during the preschool years.
HSRC. Harvard Educational Review 53:
Macdonald CA. 1990b. School-based 165–189.
Southern African Linguistics and Applied Language Studies 2002. 20: 91–103 103

Souter CW, Archer M & Rochford K. 1992. mic literacy in a second language: South
Literal and inferential reading comprehen- Africa’s educational predicament. South
sion skills. South African Journal of African Journal for Higher Education 13:
Higher Education 6(3): 30–40. 205–212.
Spear-Sterling L & Sternberg RJ. 1996. Off Webb V. 1999. Language study and Language
Track: When Poor Readers Become use in South African schools: A view from
‘Learning Disabled’. Boulder, Co: Westview the politics of language. Paper delivered at
Press. the NAETE (National Association of
Stanovich KE. 1986. Matthew effects in read- Educators of Teachers of English) 15th
ing: Some consequences of individual dif- Annual Conference, SACTE, Pretoria.
ferences in the acquisition of literacy. Wells G. 1986. The Meaning Makers: Children
Reading Research Quarterly 21: Learning Language and Using Language to
360–406. Learn. London: Hodder & Stoughton.
Strauss PR. 1995. Procedural Knowledge of Wood T. 1997. Revisiting language in educa-
ESL Readers in Decoding Expository Text. tion. South African Journal for Higher
Doctor Educationis thesis, Rand Afrikaans Education 10: 41–47.
University, Johannesburg, South Africa. Yuill N & Oakhill J. 1991. Children’s Problems
Sunday Times, 16 July 2000. in Text Comprehension. Cambridge:
Tuckman BW. 1988. Conducting Educational Cambridge University Press.
Research. New York: Harcourt Brace
Jovanovich.
Von Gruenewaldt JT. 1999. Achieving acade-