A Neurologist’s Overview of
MARTHA BRIDGE DENCKLA
Department of Neurology
Johns Hopkins Universiq
School of Medicine
Baltimore, Maryland 21218
“Developmental dyslexia” is not a synonym for discrepancy-based reading dis-
ability, which is the current legal-educational and DSM-111-R construct.] Neuro-
logical diagnosis demands exclusions (ruling out causes considered environmental
and emotional) and seeks diagnostic confirmation of a syndrome with inclusions
(cognitive deficit, developmental and family history of related deficits). “Neurolog-
ical” does not imply disease or categorical disorder; it implies some dimensional
difference within a module of brain function. Standard IQ scales are not the only
means of demonstrating normal intellectual potential; there are convincing profiles
based upon preserved neuropsychological domains that demonstrate healthy cogni-
Neurological diagnosis respects the distinction between “dyslexia-pure” and
“dyslexia-plus,”* the major contributor to the latter, dyslexia-plus, being subclinical
aspects of “attention deficit hyperactivity disorder (ADHD),” a developmental dis-
order category frequently co-morbid with dyslexia.’ Some aspects of ADHD add to
the cognitive profile the deficits of inattention, impulsivity, disorganization, and
output-inefficiency. Developmental dyslexia is not diagnosed in the context of a
persistent concurrent language disorder (of which dyslexia is then considered a
symptomatic component); less clear is the boundary between “dysphasia” and “dys-
lexia” when the former appears developmentally antecedent to the latter (i.e., spoken
language does not present a persistent problem at the stage in elementary school
when reading acquisition is the problem). Thus the degree of “purity,” both within
the linguistic domain and outside it, becomes an issue from a neurologist’s perspec-
tive in understanding developmental dyslexia.
Yet there are patients who may be characterized as examples of dyslexia-pure.
These patients fulfill the requirement (shared with the discrepancy-based reading
disability perspective) that their dyslexia is “specific,” that is, does not occur in the
context of globally/diffusely poor mental capacity. Usually their standard IQ scales
are at least average. They have no emotional (psychiatric) diagnoses, no environ-
mental disadvantages, and no educational deprivation. They have no past or present
spoken language impairments noticeable to any but the professional specialist (see
below), having “met their milestones” for understanding and expressing language
through speech. They have no past or present problems, related to subclinical
ADHD, with executive functions (payinghstaining attention, regulating impulsiv-
ity, or organizing their time or their space). They are reasonably good athletes, often
mechanically adept, quite often gifted in musical or graphic arts. They are socially
well-adapted with good interpersonal perceptiveness and tact. These dyslexia-pure
patients are, indeed, “unexpected reading failures,” but they do have subtle cognitive
Cognitive test batteries reveal that such dyslexia-pure patients are deficient within
the linguistic domain, most specifically with respect to the subdomain of phonol-
24 ANNALS NEW YORK ACADEMY OF SCIENCES
0gy.3 While semantic, syntactic, and pragmatic subdomains are relatively intact,
impaired phonological awareness, analysis (segmentation), and memory are demon-
strable. (There are some developmental or age-related ceilings and task-specific
variations on the theme; for example, dyslexia-pure cases may be deficient on
memorization of nonsense but competent by then on segmentation tasks.) Decoding
nonsense words often reveals the persistent weakness in the subdomain-of-interest
long after the intelligent dyslexia-pure person appears compensated to normal levels
of passage comprehension or even word recognition within reading tests4 A “core
deficit” of the phonological-linguistic domain is demonstrable and carries the most
weight in reading skill acquisition; followed over many years, the core deficit persists
despite what appears to be “outgrown” or “overcome” dyslexia. Specific testing
unmasks the “tin ear for language” (a colloquial synonym for phonological deficit).
As alluded to earlier, observation by a trained professional may reveal stigmata of
the “tin ear for language” in the young preschooler who is “protodyslexic,” but the
average parent, teacher, or pediatrician may not make note of the speech-sound
substitutions and mis-sequencings of these children. Even when, after referral for
their unexpectedly poor acquisition of reading skills, such children are observed to
make speech-sound errors in their conversational or narrative speech, most people
in their lives continue to regard them as “highly verbal” and successful commun-
icators. Their mixing up of words (“mord-wixing”) and/or malapropisms, if noted,
are considered amusing but of trivial communicative consequence.
Dyslexia-pure cases are not only slow, month-by-month and year-by-year, in
acquiring the skills of reading, they remain lifelong slow readers even after formal
(untimed) testing shows adequate word-recognition and even better passage com-
prehension. To a neurologist, such slowness bespeaks utilization of compensatory
circuitry, a route comprised of local or out-of-the-way detours when a direct express
connection is not available. Instead of making speech out of print, the true dyslexia-
pure person presumably uses orthographic features, semantics, and global context in
the compensatory rerouting by which she or he reads. However speculative this
“neurologizing” may appear, there is much evidence for the fact of persistent slow-
ness in the act of reading, even when formal scores of reading skiil are satisfactory
on standardized untimed tests, in those intelligent compensated dyslexia-pure adults
who as children were slow to acquire the basic reading skills.5
Those who compensate are attentive, output-efficient, strategic learners. Less
well compensated are those who lack such strengths, often (but not always) fully
eligible for the ADHD diagnosis. Rapid automatized naming, which is a kind of
verbal coding task requiring focused attention and output-efficiency with word
retrieval, prognosticates longitudinally who will compensate for phonological weak-
Thus, even if we agree to assume that dyslexia-plus is just as “neurological” as
dyslexia-pure, we confront a multivariate (or multidimensional) quantitative model
of developmental dyslexia wherein at any given moment in a lifetime, a different mix
of linguistic and executive function variables (dimensions) determines a reading
Not all dimensions of neuropsychological function contribute as much to the
multivariate quantitative determinants of dyslexia as do the linguistic and executive-
functional; visual processes, although involved, are not as salient or persistent.
Whether visual processes add on, albeit with lesser weight, to the total burden of the
poor reader, might well be explored by extension of the multivariate model. The
problem has been vagueness with which relevant visual processes have been de-
scribed. Even some of the best researchers have lumped all manner of visually based
tasks (like all of the standard Performance IQ subtests, which are very clearly
DENCKLA: DEVELOPMENTAL DYSLEXIA 25
heterogeneous if task-analyzed for what specific visual competence each requires).
Fractionation of language has not yet been matched in sophistication by fractionation
of visual processes, at least in the published research on dyslexic mechanisms. Then
there is the confounding of “visual” with “spatial” functions, often eliding the
distinction that not all spatial tasks are visual and, even more important, certainly not
all visual tasks are spatial. So it is inappropriate to conclude that dyslexia has nothing
to do with any visual dysfunction, in spite of the fact that a great many visually based
tasks have been shown to be areas of strength among dyslexic populations. To
reiterate: because some of these visually based strengths may be in the spatial domain
irrelevant to reading, and because some others may be so complex that deficient
subskills may have been masked with “success,” it is premature to dismiss all
possibility of some visual deficit contributing to dyslexia. One need not set up a
categorical “straw man” called the visual subtype. Rather, adhering to the multi-
dimensional quantitative view that brains may have mixed weaknesses in various
proportions and combinations of modular systems, there may be a discrete visual
factor in the equation. A hint that this may be so, in a limited developmental manner,
comes from clinical experience of the poor copy-forms scores (“visual-motor in-
tegration”) of many young (less than eleven year old) dyslexic children, even the
“pure” type; in their teens, such children make at least acceptable scores on such tests
without any training. My own research on map-walking yielded similar results (poor
below eleven years, superior after that). Since these visually presented tasks require
organization and production of motor output, however, the locus of deficit (or
“outgrowing” the deficit) is still to be determined. Perhaps the oculomotor system
plays a mediating role. Perhaps, as is suggested by their anterior brain width on MRI
scans, even pure dyslexics are subtly deficient in executive function.6
The problem for neurologically oriented research, compounded by the longi-
tudinal aspects of development, is how to compromise between purification and
generalization. If we insist on the detailed study of the relatively rare (yet not
impossible) pure dyslexic case, we may never see the brain of that exquisitely
explored case. Clinical samples are well known to suffer from more co-morbidities
than epidemiologic samples (unless, like this writer, you draw referrals from highly
sophisticated families who worry about any child not getting straight A grades).
Clinical samples, however, do submit to more extensive and intensive study than do
epidemiological samples (the “more” about whom we tend to know “less”). For any
type of sample there is the problem of longitudinal follow-up, environmental and/or
psychosocial instability/stress, and differential therapeutic interventions over time.
Correlative anatomic and physiologic neuroimaging is obtained in even tinier sam-
ples and by as-yet-unstandardized methods. No wonder studies “do not replicate” or
generalize each other.
With all these issues, it is amazing how much consensus exists. Clinical and
epidemiological samples have cross-validated; genetics has, while failing to pin
down the gene, provided further confirmatory evidence with respect to phonological
and executive functional domains. As long as each case study, clinical study, family
study, or epidemiological study describes its subjects and its terminology clearly and
operationally, avoiding vague descriptors and grandiose constructs, there is a good
chance that we will continue to refine a neurologically based profile of the differ-
ences underlying developmental dyslexia.’
Finally, although this is primarily a clinically based overview of what is meant
by the diagnostic label of developmental dyslexia, I would like to add two recently
acquired perspectives that have resulted from the research of the Learning Dis-
abilities Research Center of which I am Principal Investigator. The work summarized
here comes from the progress report of my esteemed colleague, Dr. Frank Vellutino,
26 ANNALS NEW YORK ACADEMY OF SCIENCES
who is Principal Investigator of our Project Four (school-based and “in the field”).
These two perspectives are on the (1) issues of gender ratios as relevant to the
biology of dyslexia, and ( 2 ) nature/nurture interactions in the production of the
reading disability phenotype.8
1. In kindergarten (KG) children tested with an extensive battery of cognitive,
linguistic, visuospatial, prereading, and early arithmetic skills, gender differ-
ences do emerge. That is, 65 percent of the KG children whose reading
readiness skills were in the lowest 10 percent of the distribution were boys,
while only 41 percent of those in the top 10 percent were boys. For general
verbal ability, gender is evenly distributed in the lowest 10 percent, but girls
are somewhat over represented in the top 10 percent. For early arithmetical
abilities, there were twice as many boys as girls in both the lowest 10 percent
and highest 10 percent of the arithmetic distribution. Thus, at this early age
when cultural factors (although not absent) are minimal, gender differences
that run in opposite directions for prereading and arithmetic emerge despite no
gender disadvantage in verbal ability.
2. Instructional differences: Vellutino and co-investigators have data on first
grade classrooms in the same geographic areahystem during the past two
academic years (1990-1991 and 1991-1992): they find that the proportion of
the school day spent on language-arts (reading-related) instruction ranges
from 15 to 41 percent, while the type of relative emphasis (code-versus-
meaning-oriented) within that time allocation varies from 1 to 45 percent on
code and from 16 to 75 percent on meaning.
This difference in reading “nurture” is bound to interact with factors of “nature”
to produce, ultimately, the achievement level that enters into the calculation of
discrepancy-based reading disability. Without factoring in the instructional vari-
ables, equating reading disability with the presumed biologic entity “dyslexia” is
doomed to produce confusion.
1. AMERICAN PSYCHIATRIC ASSOCIATION. Diagnostic and Statistical Manual of Mental
Disorders, 3d ed. (Rev.). American Psychiatric Association. Washington, D.C.
2. HUGHES,R. & M. B. DENCKLA. Outline of a pilot study of electroencephalographic
correlates of dyslexia. In Dyslexia: A n appraisal of Current Knowledge, A. L. Benton and
D. Pearl, Eds.:112-122. University Press. New York.
3. VELLuTiNo, F. R., D. M. SCANLON S. TANZMAN. Bridging the gap between
& M. 1991.
cognitive and neuropsychological conceptualizations of reading disability. Learn. Indi-
vid. Differ. 3:181-203.
4 LYON. . R. 1992. Research in Learning Disabilities: Research Directions. Report prepared
for the National Institute of Child Health and Human Development. National Institutes
of Health. Bethesda, Md.
5. DENCKLA, B. & J. M. RUMSEY.
M. 1992. Developmental dyslexia. In Diseases of the
Nervous System, 3d ed., Vol. 1, Clinical Neurobiology, A. K. Asbury, G . M. McKhann,
and W. I. McDonald, Eds.: 636-645.
6. HYND, . W., e t a / . 1990. Brain morphology in developmental dyslexia and attention deficit
disorderhyperactivity. Arch. Neurol. 47:919-926.
7. PENNiNGToN, B. F. 1991. Dyslexia and other developmental language disorders. I n Diag-
nosing Learning Disorders, chap. 4:4.5-81. Guilford Press. New York.
8. VELLUTINO,R. & D. M. SCANLON. Intervention-resistant reading disability. Progr.
Rep, Project IV. P.50 HD2.5806 Neurodevelopmental Pathways to Learning Disabilities.
M. B. Denckla, PI.