Dogs, Tricks, and some Grand Claims
Despite the old adage about dogs and tricks, nearly thirty years of teaching representational, skill-based drawing and painting has given me little reason to believe that age is a make-or-break factor in either learning or the performance that may follow. Students differ enormously, of course, but those differences do not sort themselves neatly by age. Some younger students learn quickly and well–some do not. Some older students arrive with reinforced habits and deeply embedded assumptions that must be examined and revised. Yet their more extensive life experience may also bring patience, discipline, accumulated knowledge, and a clearer sense of purpose. These are all qualities that can matter just as much as speed or youthful flexibility. Again and again, I have watched eager learners from eight to eighty acquire abilities they had once assumed were beyond their reach. For a teacher, that is, of course, deeply fulfilling. However, that reality has remained in productive tension with something I first encountered as a student in college.
I entered the commercial art program at a local college in the early 1990s, at a time when skill-based representational art was beginning to re-emerge after a long period of marginalization within much of higher art education. I was fortunate in that my instructors and mentors were highly skilled in their respective disciplines and, more importantly, equally skilled at transmitting both knowledge and practice to their students.
Around the department, teachers and students circulated the classical realist journals associated with the Boston ateliers, along with a book that many came to regard as something close to a manifesto for the realist revival: R. H. Ives Gammell’s Twilight of Painting. First published in 1946, the book argued that the traditional craft of painting had deteriorated, that the disciplines once transmitted through ateliers and academies had been abandoned, and that modern painting had suffered as a result. Contemporary and later readers recognized the book’s importance to the classical realist movement, even when they disputed its conclusions. A 1948 review in The Journal of Aesthetics and Art Criticism treated it as a serious contribution to debates over art education, while later realist writers described it as foundational to the recovery of traditional studio practice.
Among Gammell’s many claims, one passage about inborn ability and the age of the learner remained especially vivid to me. Over the decades, as I taught students of widely varying ages, it returned again and again to the forefront of my thinking.
On page 36 of Twilight of Painting, Gammell writes:
“Therein lies the overwhelming importance to a painter of a sound basic training received in early life. Of all the procedures, devices, methods, and special abilities required by an artist to transform his initial conception into a picture likely to prove a viable work of art, the ability to represent the appearance of whatever he may select is the most important and the most difficult to acquire. This ability can, in fact, only be mastered by a person with an exceptional inborn aptitude for painting. But even a person so gifted will be able to set down the shapes and colors of things with accuracy and distinction only after a long training under a master himself proficient in the art. It has been abundantly proved that such a training is only effective during an artist’s formative years. After his early twenties, his perception of form ceases to be sufficiently educable. The great painters of the Renaissance and of the seventeenth century started their apprenticeships around fourteen or fifteen at the latest. I do not know of a case of a draftsman approaching the first rank whose training started later than eighteen. The errors and omissions of a painter’s early training are irreparable.”
It is difficult to read that passage without pausing over its certainty. “Only” a person with exceptional inborn aptitude can master representation? “Abundantly proved” that training is effective only during formative years? Apprenticeships by fourteen or fifteen to even have a chance to be “great”? Are the early twenties the cutoff for form perception to be sufficiently educable? Are all of these early errors irreparable?
These are not anything close to modest observations. They are sweeping grand claims about talent, development, education, and the limits of the adult mind. Yet some of the key terms remain strikingly undefined. What counts as “exceptional inborn aptitude”? Which years qualify as “formative”? What evidence could justify the claim that perception ceases to be educable after the early twenties? And does the historical fact that many Renaissance artists entered workshops in adolescence prove a biological necessity—or merely describe the educational and social structures of their time?
That is the question that I aim to examine with this essay.
Was Gammell identifying a genuine limit in human development? Is he mistaking a historical pattern of apprenticeship for a neurological law? Or is he giving scientific-sounding authority to a cultural myth about youth, talent, and artistic legitimacy? The claims are large enough to deserve more than agreement or dismissal. They deserve evidence.
To be Fair…
Before addressing these claims one by one, I want to take a moment to explicitly state that this writing is not about Gammell the person or the artist, but rather about these ideas as they are presented in the book. I believe they SHOULD be addressed, as they continue to permeate some circles to this day, but I do not want to suggest in the slightest that Gammell should be treated as though he invented the idea of fixed aptitude or a narrow formative window in isolation.
Twilight of Painting appeared in 1946, in an intellectual climate very different from our own. Ideas about heredity, aptitude, maturation, and age-linked development carried considerable authority in psychology and education. Intelligence testing had become influential. “Talent” was often treated as a relatively stable endowment, and educational systems routinely sorted learners according to presumed ability. Development itself was commonly discussed in terms of stages, norms, and formative periods, and within that setting, the idea that certain capacities had to be established early would not have sounded as extraordinary as it does today.
It is also important to note that Gammell was writing from within a threatened artistic tradition. He believed that representational painting depended on a demanding chain of transmission from master to pupil: drawing, observation, correction, repetition, and the gradual acquisition of judgment under the supervision of an experienced practitioner. In his view, that chain had been weakened by modernism and by changes in institutional art education. Much to his credit, later accounts of Gammell and his students describe him as a defender of a craft tradition he believed was in danger of disappearing. Seen from within that worldview, the apparent anxiety in some of the writing becomes easier to understand. If representational skill must be built through years of supervised practice, if mistakes become deeply ingrained, and if competent teachers are becoming scarce, then the loss of traditional training would appear not merely regrettable but catastrophic. His claim that early errors are “irreparable” serves this larger argument, but it also raises the stakes. A broken educational tradition does not simply delay the student; it may also permanently damage the future of skill-based painting and drawing.
His historical examples would also have seemed persuasive. Many celebrated painters of the Renaissance and the seventeenth century did begin apprenticeships in adolescence. Gammell could point to that pattern and see confirmation that serious training had always begun early. What he did not sufficiently separate was the biology of learning from the social organization of apprenticeship. Young people entered workshops early because that was how trades, households, guilds, and vocational education were structured. So while the historical pattern was real, the meaning in the context is far less certain.
The scientific picture of adult learning in Gammell’s own time was not one-sided. Psychologists had begun to study adult learning directly. American psychologist Edward Thorndike’s work in the late 1920s was influential precisely because it challenged the assumption that education properly belonged only to the young. Thorndike reported age-related differences and a decline in certain measures after early adulthood, but his broader conclusion was that adults remained capable of learning. Later histories of adult education treat this work as foundational to the recognition that learning continues well beyond youth.
This matters because Gammell’s statement was not simply the best available scientific conclusion of 1946. The period contained competing ideas. One current emphasized fixed aptitude, maturation, and decline, while another emphasized continued educability and the growing legitimacy of adult learning. Gammell’s passage aligns strongly with the first current, but I do not believe that we know whether he drew on any specific scientific source. The fairest interpretation may be that Gammell combined several kinds of evidence that appeared mutually reinforcing to him: the early apprenticeships of canonical painters, his own experience of students whose habits had become difficult to correct, a cultural belief in exceptional inborn talent, and a broader educational language of formative development. However, it is also fair to state that from those observations, he drew a conclusion that exceeded what they could actually establish. So it is not fair to say these statements are foolish, but rather, historically situated.
So I offer for your consideration that we grant the strongest version of Gammell’s position: he was defending a demanding craft at a moment when its institutions seemed to him to be collapsing, and he was writing within a culture that often treated aptitude and development as more fixed than we do today. But fairness also requires the next question:
Even in the context of his own time, were those claims actually justified?
That is where our historical framing must give way to the evidence.
Historical Pattern Is Not Biological Proof
Since we are already in the territory of history, let us begin by taking Gammell’s historical observations seriously. Many of the canonical painters of the Renaissance and seventeenth century did begin their training young, often as adolescents in workshops governed by guilds, masters, family arrangements, and long-established systems of apprenticeship. In that sense, his statement is not invented. Early entry into a workshop was simply a normal feature of the period’s professional structure. Art education was embedded in labor and household organization rather than separated from them, as modern schooling often is. (Bignamini & Castelnuovo, 1998; Efland, 1983)
But a historical pattern of early entry into such systems is not, by itself, evidence of a biological cutoff. That distinction is easy to miss because the facts appear to line up so neatly. Great painters often began young. They trained for years and became highly accomplished. It is therefore tempting to infer that beginning young was neurologically necessary. Yet the historical record cannot tell us that on its own. It tells us when artists entered the systems available to them, but it does not reveal what might have happened had equally motivated adults been given comparable access, instruction, time, and social permission to begin later.
Apprenticeships were also shaped by practical realities. Families placed young people into trades. Workshops depended on inexpensive labor, while guilds regulated entry, production, and advancement. Training often began at the age when a young person was expected to leave childhood and enter the arena of work. An apprentice did not merely study drawing and painting; they prepared materials, copied, cleaned, observed, assisted, and gradually took on more complex tasks. The age of entry reflected the organization of the profession at least as much as the development of the brain.
This is the central problem in Gammell’s reasoning:
A historical pattern of early training is treated as evidence that later learning is biologically problematic. Let’s parse the propositions interwoven so that they can easily collapse into one another:
| Proposition | Assessment |
| Many canonical artists began young | Historically evident in many cases |
| Starting young provides more years of practice | Clearly true |
| Early training may improve the odds of elite attainment | Plausible |
| Adult perception becomes uneducable | Not supported |
The first three propositions can all be true without the fourth being true.
It is absolutely true that beginning at fourteen gives a learner more time than beginning at forty. It may allow more years of supervised practice, more repetitions and iterations, more opportunities to correct errors, and more time for certain skills to become integrated into identity and daily life. In highly competitive fields, those advantages may matter enormously, and with such considerations in mind, early training may well increase the probability of exceptional attainment.
But youth does not possess every advantage in this context. An older learner may arrive with less time and more deeply established habits, but they may also bring resources largely unavailable to the adolescent beginner: accumulated knowledge, broader experience, stronger self-direction, greater patience, and a clearer sense of why the skill matters. Mature learners may be better able to organize practice around long-term goals, connect unfamiliar ideas to existing knowledge, and tolerate the slow correction of errors because they have chosen the undertaking deliberately rather than inherited it as part of childhood training.
That said, it would be equally misleading to over-romanticize prior knowledge. Yes, it can absolutely support increased comprehensive learning, but it can also interfere with it. Established concepts may help a learner recognize structure, analogy, and meaning; they may also encourage premature conclusions or make old habits more resistant to revision. The fairest summary may be that the adult learner can bring a richer foundation, but often a more complicated one.
That difference matters. The comparison is not simply between a highly plastic young brain and a potentially more rigid older brain. It is between two learners who possess different combinations of flexibility, experience, time, knowledge, motivation, and habit. Again, early training may increase the probability of certain forms of attainment, but increased probability is not the same as necessity. Nor does the prominence of early starters in the historical canon prove that later starters were incapable. The canon is not a neutral sample of human potential. It reflects who had access to training, who could afford it, who was admitted to workshops and academies, who received patronage, whose work was preserved, and whose careers later counted as evidence of greatness. Women, laborers, outsiders, and adults without institutional access were often excluded before their abilities could ever be tested. The absence of late beginners from the historical record may therefore reveal as much about opportunity and selection as about capacity. Histories of the canon repeatedly show how access to apprenticeship, academies, patronage, and professional recognition shaped who could become visible as an artist in the first place. (Broude & Garrard, 1987)
There is also a survivorship bias that should be mentioned. We remember the artists who entered established systems, accumulated years of practice, found patrons, and produced work that institutions preserved. We do not have a comparable record of those who began outside those systems, trained informally, or never gained access to the professional networks through which artistic reputations were built.
For that reason, Gammell’s statement— “I do not know of a case of a draftsman approaching the first rank whose training started later than eighteen” —has limited evidentiary force. It tells us something about the cases known to Gammell, but it does not establish the limits of adult learning.
This also does not mean that age is irrelevant, since that would simply replace one overstatement with another. Early training may confer real advantages in practice time, habit formation, fluency, and eventual level of attainment, so the more careful conclusion is that age affects the conditions under which skill develops, but it does not follow that age determines whether development remains possible.
The distinction can be stated plainly:
Early opportunity and biological necessity are not the same thing.
Now, before we can decide whether Gammell was right about the limits of adult perception, the historical record must be treated as insufficient proof. The next question belongs not to biography or apprenticeship history, but to the science of learning itself.
Fact and Fiction
Before jumping right to individual claims, it would seem helpful if we established a more careful framework. Discussions of age and learning often collapse into one of two extremes. In the first, the brain is treated as essentially finished after youth, leaving the adult learner to work around a system that can no longer change meaningfully. In the second, neuroplasticity is invoked as proof that anyone can learn anything at any age, provided they try hard enough.
Neither position is scientifically adequate. The evidence supports a more demanding conclusion: the brain remains capable of change across the lifespan, but the conditions, rate, mechanisms, and likely extent of that change are not identical at every age. Adult learning is real. So are age-related differences. The challenge is to describe both without turning either into a slogan.
So how might we address the myth that the brain stops changing after youth? The most direct response is that it does not. Adult brains remain quite plastic. They continue to change in response to experience, practice, injury, environment, and learning. In neuroscience, plasticity refers broadly to the nervous system’s capacity to alter its organization or function. Depending on the level being studied, that change may involve synaptic strength, patterns of neural recruitment, cortical representation, connectivity, sensory weighting, motor commands, or the efficiency of task performance. This matters because drawing and painting are not purely intellectual activities. They require repeated coordination among perception, attention, memory, decision-making, and movement. Learning to judge an angle more usefully, organize a value structure, subtly vary an edge, or coordinate hand movement with visual information involves systems that absolutely remain trainable in adulthood.
Research on motor learning is especially clear on this point. Adults can acquire new movement sequences, refine existing actions, improve prediction and error correction, and become more efficient through practice. Motor training has been associated with experience-specific functional changes across motor networks, while reviews of skill learning describe plastic changes in cortical and subcortical systems that support acquisition and retention. These mechanisms do not disappear when adolescence ends. (Adkins et al., 2006; Dayan & Cohen, 2011)
The same general conclusion appears in research on cognitive training and aging. Older adults often improve on trained tasks, and those improvements may be accompanied by changes in patterns of brain activation or recruitment. Reviews of cognitive aging, therefore, reject the idea of a completely fixed older brain. They describe an aging nervous system that continues to adapt, although sometimes through recruitment patterns different from those seen in younger adults. (Lustig et al., 2009; Hertzog et al., 2008)
The important word here is adapt.
Learning-related brain change does not always mean that some new structure visibly appears, nor does increased activation automatically mean improvement. A brain-imaging difference may reflect greater efficiency, greater effort, compensation, a change in strategy, or some combination of these. It is therefore safer to say that learning can alter neural function and organization than to claim that every improvement produces a simple, easily interpreted rewiring of the brain. This distinction protects the argument from a common misuse of neuroscience. A student can improve without a scan proving why, and a scan can show change without establishing that the change caused the improvement. Behavioral evidence—what a learner can now perceive, remember, or do—is what will remain central for our purposes here..
It is also important to note that plasticity changes with age. Although it is very true that plasticity persists, it does not remain constant. Some forms of learning are faster or more efficient earlier in life. Younger nervous systems may show greater sensitivity to certain kinds of input, greater flexibility in some circuits, faster processing, and fewer entrenched habits competing with new learning. Age can also affect reaction time, sensory acuity, working memory, motor variability, fatigue, and the speed with which a new skill becomes stable. Therefore, we may find that older adults require more practice, more explicit instruction, a different pacing, or more recovery time to achieve the same degree of improvement on some tasks. Studies of motor learning generally find that older adults continue to improve, but may show slower acquisition, greater variability, or weaker performance when tasks place heavy demands on speed, sequencing, divided attention, or coordination. (Voelcker-Rehage, 2008; Cai et al., 2014)
But reduced speed is not the same as lost capacity. A slower rate of acquisition does not demonstrate that the system is no longer educable. Nor does a lower average performance in an older group establish the limit of a particular older individual. Age-group findings describe tendencies across populations; they do not function as expiration dates. This is especially important in a studio setting. A learner may take longer to coordinate a new brush movement, discriminate a subtle interval, or inhibit an old drawing habit. Yet with appropriately structured practice, performance may still improve substantially. The practical question is often not whether change is possible, but what conditions make that change more likely.
We also have to be clear that plasticity is not “magic”. Unfortunately, the modern enthusiasm for neuroplasticity has produced its own mythology. The phrase is sometimes used to suggest that the brain can be reshaped without meaningful constraint, as though effort alone guarantees any desired outcome. That is not what the evidence shows.
Plasticity is selective, task-dependent, and constrained. It is shaped by what is practiced, how often it is practiced, how difficult the task is, whether feedback is available, whether the learner can detect and correct errors, and whether the training continues long enough to stabilize improvement. It is also affected by health, stress, fatigue, sensory loss, medication, injury, sleep, motivation, and prior experience. Sleep deserves particular attention because newly acquired skills are not consolidated only during practice. Memory and motor learning continue to be processed after practice, and age-related changes in sleep may affect that consolidation. Poor sleep does not make learning impossible, but it can reduce the effectiveness of training and recovery. (Mander et al., 2017)
Prior experience also matters in two directions. Existing skills can provide useful scaffolding, allowing new information to be connected to familiar concepts or motor patterns. But previous learning can also interfere when an old response competes with a new one. The adult learner’s nervous system is therefore not simply less plastic; it is more historically committed. It already contains many practiced solutions, expectations, and habits. That can be an obstacle, but it can also be an asset.
Another qualification is necessary: Improvement is usually specific before it is general. Training commonly improves the trained task and closely related tasks, while broader transfer is less reliable. A learner who practices ahep replication may improve at shape replication. A learner who repeatedly organizes values may become better at organizing values. A learner who practices controlled mark-making may gain more reliable control. These are meaningful changes. But they do not automatically imply broad gains in intelligence, memory, creativity, or every other cognitive function. Reviews of cognitive training repeatedly caution that near transfer is more common than far transfer. In plain language, practice tends to improve what is practiced most directly. Claims that a single activity broadly transforms the entire brain should therefore be treated with caution. (Green & Bavelier, 2008; Hertzog et al., 2008)
For an article about artistic learning, that is not a disappointment. Representational drawing and painting do not need to raise general intelligence in order to count as serious adult learning. If practice improves the specific perceptual, motor, and decision-making abilities required by the art, that is sufficient.
So what might all this mean for a claim like Gammell’s, in which, after the early twenties, a painter’s “perception of form ceases to be sufficiently educable.” Fortunately, modern evidence does not support such a categorical statement. As I have explained, while it is true that adult visual, cognitive, and motor systems remain responsive to training, their responsiveness may differ from that of younger systems. Improvement may be slower, more variable, or more dependent on deliberate structure. Some forms of plasticity may be reduced, and some habits may be harder to revise. But none of that, and I mean none of it, amounts to the cessation of educability.
The scientifically responsible conclusion is more modest: The capacity for highly effective change persists, while the conditions and efficiency of change may alter with age.
That is very different from saying that the adult brain can do anything, and it is also very different from saying that it can no longer learn. The adult brain is not fixed, but neither is it developmentally identical to the child’s brain. That distinction will become increasingly important as we turn from the general question of plasticity to the more specific claims about critical periods, perception, motor skill, and the supposed irreparability of early errors.
So what are some relevant factors that actually change with age?
Once the oversimplifications and slogans are stripped away, the picture that remains is neither simple decline nor simple continuity. Age changes the learner, but it does not change every capacity in the same direction, at the same rate, or to the same degree. Some abilities associated with speed, rapid adaptation, working memory, and sensory-motor efficiency often favor younger adults on average. Other resources, such as knowledge, strategy, selectivity, self-regulation, and accumulated context, may become more readily available with experience. The result is not a clean exchange in which youth loses exactly what adulthood gains. It is more appropriate to view this as a change in the learner’s overall profile.
A useful summary looks something like this:
| Dimension | Typical youthful advantage | Possible adult advantage |
| Processing speed | Faster responses and quicker uptake | More strategic pacing and selective effort |
| Motor adaptation | Often quicker adjustment to novel demands | More deliberate correction and monitoring |
| Practice horizon | More years available for accumulation | Stronger goal selection and commitment |
| Prior knowledge | Fewer competing assumptions | Richer schemas, analogies, and associations |
| Habit strength | Fewer entrenched habits | Greater self-knowledge about recurring habits |
| Motivation | Often externally structured by school or family | More personally chosen and meaning-driven |
| Error tolerance | Sometimes less self-conscious about failure | Sometimes more disciplined and persistent |
| Interpretation | Less accumulated cultural and personal context | Greater experiential and conceptual depth |
The terms “typical” and “possible” matter here. These are not guarantees. A younger learner may be unusually reflective, disciplined, and richly informed. An older learner may be impatient, poorly motivated, or unaware of entrenched habits. Age-group differences describe tendencies, not destinies.
Processing speed is one of the clearest examples. On many laboratory tasks, younger adults respond more quickly and handle rapid information more efficiently. That can matter when a task places a premium on speed, divided attention, or the immediate coordination of multiple demands. But, as some of you are likely already aware, studio learning is usually a race. An adult learner may compensate by slowing the task down, organizing it more deliberately, and allocating attention more selectively. Research on aging increasingly emphasizes this strategic selectivity: older adults may invest effort differently, especially when a task is meaningful or clearly tied to a goal. (Hess, 2014; Swirsky & Spaniol, 2019)
Motor learning shows a similar pattern. Younger adults often adapt more rapidly to new movement demands. Older adults may require more repetitions, more feedback, or more recovery, especially when the task depends heavily on speed or coordination. Yet they continue to improve, and slower adaptation does not, in any way, imply the absence of learning. In a drawing or painting context, rapid acquisition may matter less than the ability to identify an error, isolate its cause, and work deliberately toward correction.
Practice horizon also deserves careful treatment. The young learner has the obvious advantage of time. Beginning at fourteen creates more years in which skill can accumulate than beginning at fifty. That advantage is real and should not be inappropriately minimized. But available years and usable practice are not identical. A mature learner may bring a clearer sense of priorities, greater control over practice conditions, and a more deliberate commitment to a chosen goal. Thus, while the younger learner may have more time, the adult may use limited time with greater selectivity.
Prior knowledge may be the most consequential difference for the argument developed here. Because younger learners possess fewer established conceptual structures, they may be less constrained by familiar procedures when encountering novel tasks. Adults, by contrast, draw on broader stores of semantic knowledge and more extensive associative networks. These resources can significantly expand the potential for meaningful learning by enabling the integration of new information with a greater array of existing knowledge. This can be a powerful advantage. A mature learner may grasp the concept of visual hierarchy through prior experience in design, music, language, or engineering, or through observation of the natural world. They may understand proportional relationships by analogy, connect value structure to broader ideas of organization, or interpret pictorial choices through historical and cultural knowledge. But, to be fair, prior knowledge is not automatically helpful. It can sometimes be misleading, distort perception, or lead to a resistance to correction. A student who “knows” what an eye, tree, or hand looks like may rely on a stored symbol rather than attend to the particular form in front of them (schematic substitution/conceptual contamination). The same knowledge network that provides a greater landscape for connection can also produce interference.
Habit strength follows the same double pattern. Younger learners may have fewer entrenched motor and perceptual habits, while adults may reasonably have more. Yet mature learners may also possess greater self-knowledge. This means that they may recognize how they respond to frustration, how long they can concentrate effectively, what kinds of feedback help them, and which errors they repeat. Such metacognitive awareness can support more deliberate practice, though it, too, varies widely among individuals. (Price, Hertzog, & Dunlosky, 2010; Sheffler et al., 2022)
Motivation also changes in form. Younger learners often work within externally structured systems: school schedules, parental expectations, grades, assigned exercises, and institutional deadlines. Adults, on the other hand, are far more likely to enter a training or educational scenario voluntarily. That does not mean they are always more motivated, but their motivation may be more personally chosen and more tightly connected to identity, meaning, or long-deferred goals. Research on cognitive aging suggests that motivation and perceived value can substantially influence how older adults allocate attention and effort. (Hess, 2014)
Error tolerance also resists any neat division by age. Some younger learners are relatively unselfconscious and willing to produce imperfect work, while others are intensely sensitive to judgment. Adults can be equally varied: some are constrained by perfectionism or by the fear that mistakes represent wasted time, while others bring patience, humility, and the discipline to remain with a problem through repeated failure. Pedagogically, the more important variable is not age itself, but the learner’s relationship to error.
Finally, interpretation, closely related to prior knowledge, may be the area in which more mature learners bring the most obvious depth. Representation is not only a matter of fine motor control or trained perceptual practices; it also depends on judgment, emphasis, selection, memory, and the assignment of meaning. A learner with decades of experience may bring a broader field of reference to tasks of visual interpretation and image-making. That accumulated context does not guarantee “better” representation (whatever that may truly mean), but it can enrich what the learner notices, values, connects, and ultimately chooses to communicate.
For that reason, the central question is not whether youth or age is better. It is:
Which capacities does the task require, which capacities does the learner currently possess, and how can instruction connect the two?
That is a more useful question because skill-based visual arts practices are not a single capacity. A student may be strong in observation but weak in motor control, rich in conceptual understanding but poor at navigating error correction, highly motivated but inefficient in practice, or technically competent but limited in communicating intent. Age may influence each of these components differently. The practical consequence is that instruction should not be organized around a generic idea of the “young” or “old” brain. It should be organized around the actual learner.
A younger student may need structure, patience, and conceptual depth. An older student may need help loosening a habitual solution, reducing unnecessary cognitive load, or breaking a complex task into slower and more manageable parts. Both may need repetition, feedback, and time, while the proportions may differ significantly.
The Mature Learner’s Double Inheritance
Decades of living, working, observing, solving problems, and forming judgments create a large store of knowledge and potential for association. Aside from a limited class of non-associative processes, such as habituation and sensitization, most new learning depends on forming, modifying, or integrating associations with what the learner already knows, remembers, expects, and has experienced. Adults may therefore bring stronger strategies, greater self-knowledge, more deliberate goals, and a clearer sense of why the work matters
But as I’ve mentioned earlier, the same history that supplies these resources can also create interference. Established knowledge can become expectations, familiar interpretations can harden into assumptions, and efficient habits can become automatic responses that go unnoticed. A learner who has become competent in other areas of life may find beginnerhood unusually uncomfortable, especially when difficulty threatens a long-established sense of identity. The adult may know more, but may also have more invested in already knowing.
This is why prior experience cannot be classified simply as an advantage in all cases. It functions more like an inheritance: valuable, extensive, and not always easy to manage.
| What accumulated experience can provide | What accumulated experience can complicate |
| Greater knowledge and conceptual range | Assumptions that feel like facts |
| Greater analogy and pattern recognition | Premature reliance on familiar comparisons |
| More refined strategic thinking | Reluctance to experiment |
| Self-knowledge | Self-consciousness and fear of incompetence |
| More established habits | Resistance to revising automatic responses |
| Greater interpretive depth | Overconfidence in established meanings |
| Stronger personal purpose | Greater identity investment in success |
| Greater landscape of prior experience | Interference from previously learned solutions |
The two sides are not independent. Often, they are the same capacity viewed under different conditions. Pattern recognition can slip into disadvantageous stereotyping, confidence can manifest as rigidity, and discipline can become excessive self-monitoring. Experience can seem to expand our scope of perception, but that expansion can also carry a greater tendency to get mired in expectation. This is especially relevant to observational representationalism. Students bring stored symbols, verbal labels, habitual simplifications, and cultural conventions to their visual perceptions of a face, a hand, a tree, or an interior. Those resources help them identify and organize what they see, but they can also override certain aspects of observation. The student does not draw the specific eye in front of them, but the categorical “eye” that is already established in memory (again, this is what is known as schematic substitution or conceptual contamination)
All learners must connect new learning to what is already known and revise what is already known when it interferes with accurate perception or effective action. For mature learners, however, both sides of that process may be more extensive because more knowledge, experience, and habits have accumulated. The tension can be stated simply:
The mature learner has more to connect with—and often more to overcome.
This is not a defect of adulthood. It is an intensified version of a condition shared by all learners: learning always proceeds through an existing history.
The pedagogical consequence is important. Learning is not simply a matter of adding new information to what a student already knows. Existing knowledge and habits may help a learner attend, interpret, and act more effectively, or they may narrow perception and resist correction. Good instruction helps learners recognize the difference: preserving what remains useful, revising what misdirects attention or action, and bringing automatic habits into conscious examination.
That is where metacognition (awareness and regulation of one’s own thinking and learning) becomes especially important. The adult learner’s advantage is not simply that they possess more knowledge, but that they can examine how that knowledge is used. When learners can identify the assumptions guiding a mark, the habits shaping a response, or the expectations distorting observation, prior experience becomes available for revision rather than remaining an invisible constraint. Research on adult learning and metacognition supports this broader view: prior knowledge can facilitate new learning when it is relevant and accurate, but it can impede learning when it is incomplete, inappropriate, or applied automatically. Metacognitive monitoring helps determine which is occurring. (Efklides, 2014; Birney, Beckmann, & Wood, 2012; Lovett et al., 2023). This makes the adult learner neither less educable nor automatically wiser. It makes adult learning more historically layered.
Returning to Twilight
With all of this in place, Gammell’s claims can now be reconsidered more effectively. The issue is not whether early training can make a difference in outcome–it clearly can. Nor is it whether starting young may improve the odds of attaining exceptional mastery. It may. The question is whether those observations justify the stronger conclusion that adult learners become fundamentally incapable of acquiring high-level perceptual and representational skills.
Put simply…they do not.
“After his early twenties his perception of form ceases to be sufficiently educable.”
This is the most categorical and least defensible claim in the Twilight passage. Adult perception does not cease to be educable. Perceptual discrimination, visual attention, motor coordination, and task-specific judgment remain trainable across adulthood. Age may affect some of the dynamics of learning, the amount of practice required, and the limits reached under particular conditions. Those are meaningful differences, but they are not equivalent to some magic end of educability.
The phrase “ceases to be sufficiently educable” also conceals an undefined standard. Sufficient for what? competent representation? professional work? exceptional draftsmanship? or inclusion among the canonical masters? Without a clear criterion, the claim cannot be tested as stated. It moves too easily from the reasonable observation that some capacities change with age to the far stronger assertion that adult development is no longer adequate for serious artistic achievement.
That conclusion is not supported by the broader evidence. It is a developmental possibility stated as a biological finality.
“The great painters of the Renaissance and of the seventeenth century started their apprenticeships around fourteen or fifteen at the latest.”
As a historical observation, this is, as we’ve discussed earlier, broadly plausible. Many artists entered workshops at ages that would now seem unusually young. But apprenticeship age tells us as much about the organization of artistic labor as it does about human development.
Workshop training was embedded in household economies, guild structures, family networks, and systems of occupational preparation that began in adolescence. Young apprentices did not simply receive an early neurological advantage. They also gained years of supervised practice, repeated exposure to professional standards, access to materials and commissions, and a place within institutions that controlled entry into the field. Beginning at fourteen or fifteen, therefore, had obvious practical consequences. It increased the duration and intensity of training and placed the learner inside a professional culture at an early age. But a historical pattern of early entry does not establish that later entry is biologically ineffective.
Early opportunity and biological necessity are not the same thing.
“I do not know of a case of a draftsman approaching the first rank whose training started later than eighteen.”
This is candidly framed as a statement about Gammell’s knowledge, not as the result of a systematic investigation. “I do not know of a case” may identify a pattern worth examining, but it cannot establish that no such case exists or could exist.
The phrase “first rank” also introduces a second problem. Artistic rank is not a neutral scientific category. It depends on canons, institutions, critical judgments, access to training, preservation, reputation, and historical visibility. A canon composed largely of artists who entered established systems at a young age will naturally reproduce the impression that greatness requires early entry. Those who began later, trained outside recognized institutions, worked under conditions of exclusion, or failed to enter the historical record are less likely to appear in the sample from which the conclusion is drawn.
This does not prove that late beginners commonly reach the highest levels of draftsmanship. It shows that Gammell’s evidence cannot support the universality of his claim. The absence of a known example is not evidence of impossibility, especially when the category being examined has already been shaped by access, selection, and survivorship.
“The errors and omissions of a painter’s early training are irreparable.”
This claim also goes too far. Early training can have lasting effects. Habits of mark-making, proportional judgments, value organization, spatial construction, and paint handling may become deeply established through repetition. Correcting them may require sustained attention, deliberate practice, and the temporary disruption of procedures that have become automatic.
But resistant is not irreparable.
Adults can reorganize habits, refine discrimination, adopt new strategies, and retrain coordinated actions. Some earlier patterns may never disappear completely, and some limitations may indeed remain (it depends on many factors). That is very different from saying that errors of early training cannot be repaired at all. In practice, artistic development often includes precisely this kind of revision: abandoning formulas, correcting misconceptions, restructuring procedures, and learning to notice what previous training encouraged the artist to overlook.
The stronger conclusion warranted by the evidence is therefore modest:
Early training matters because habits accumulate. It does not follow that later correction is impossible.
Gammell was right to emphasize the difficulty of representational mastery and the value of sustained instruction under a capable teacher. He was also right that early training can confer certain advantages. Where his argument fails is in converting those advantages into a fixed developmental deadline.
Nothing here proves that every late beginner can become a canonical master and no serious account of learning could promise that. Exceptional achievement depends on many interacting conditions, including (but not limited to) aptitude, health, opportunity, instruction, motivation, time, resources, social support, and sustained practice. Starting late may narrow some possibilities, especially where the remaining practice horizon is short or the desired level of attainment is extraordinarily high.
But none of that justifies saying that adult perception ceases to be educable, that no first-rate draftsman can begin late, or that early errors are beyond repair. The defensible conclusion is narrower and more useful: age changes the conditions of learning, sometimes substantially, but it does not establish a point at which serious artistic development becomes impossible—or, by age alone, even improbable.
Replacing the question
The common question—“Am I too old?”—is just poorly framed in this context. It looks to age to provide a single answer to a problem shaped by many, many variables: the skill being learned, the learner’s prior experience, the quality of instruction, the structure of practice, and the time and effort available. More useful questions in this arena might be:
- What component skill am I trying to learn?
- What prior knowledge can support it?
- What habit or expectation might help or interfere?
- What kind of instruction, feedback, and practice is most likely to produce change?
- What constraints of time, health, attention, or opportunity must be taken seriously?
These questions do not deny that age matters. They place age within a larger model of learning rather than allowing it to function as a verdict. To be clear: there is no scientific basis for treating fourteen, eighteen, or the early twenties as the expiration date for serious artistic learning. Age matters, but it does not operate as a switch that turns educability off. It changes the learner’s resources, constraints, pace, and available horizon. The mature student may learn more slowly in some respects, yet bring to the task a denser architecture of knowledge, memory, strategy, judgment, and meaning. Formal training does not ask the learner to become young again, nor does it ask for the erasure of everything already known. It asks for something more difficult and more interesting: the deliberate reorganization of an existing mind.
Resources
Primary and Historical Sources
Gammell, R. H. I. (1946). Twilight of Painting.
Thorndike, E. L., Bregman, E. O., Tilton, J. W., & Woodyard, E. (1928). Adult Learning. New York: Macmillan.
Singerman, H. (1999). Art Subjects: Making Artists in the American University. University of California Press.
Smith, P. (1996). The History of American Art Education: Learning About Art in American Schools. Greenwood Press.
Winner, E. (2022). An Uneasy Guest in the Schoolhouse: Art Education from Colonial Times to a Promising Future. Oxford University Press.
Adult Learning, Aging, and Skill Acquisition
Dayan, E., & Cohen, L. G. (2011). Neuroplasticity subserving motor skill learning. Neuron, 72(3), 443–454.
Hertzog, C., Kramer, A. F., Wilson, R. S., & Lindenberger, U. (2008). Enrichment effects on adult cognitive development: Can the functional capacity of older adults be preserved and enhanced? Psychological Science in the Public Interest, 9(1), 1–65.
Voelcker-Rehage, C. (2008). Motor-skill learning in older adults: A review of studies on age-related differences. European Review of Aging and Physical Activity, 5, 5–16.
Green, C. S., & Bavelier, D. (2008). Exercising your brain: A review of human brain plasticity and training-induced learning. Current Directions in Psychological Science, 17(1), 16–20.
Sheffler, P., Rodriguez, T. M., Cheung, C. S., & Wu, R. (2022). Cognitive and metacognitive, motivational, and resource considerations for learning new skills across the lifespan. WIREs Cognitive Science, 13, e1585.
Hess, T. M. (2014). Selective engagement of cognitive resources: Motivational influences on older adults’ cognitive functioning. Perspectives on Psychological Science, 9(4), 388–407.
Prior Knowledge, Association, and Metacognition
Shing, Y. L., & Brod, G. (2016). Effects of prior knowledge on memory: Implications for education. Mind, Brain, and Education, 10(3), 153–161.
Gilboa, A., & Marlatte, H. (2017). Neurobiology of schemas and schema-mediated memory. Trends in Cognitive Sciences, 21(8), 618–631.
Parsons, R. G. (2018). Behavioral and neural mechanisms by which prior experience impacts subsequent learning. Neurobiology of Learning and Memory, 154, 22–29.
Efklides, A. (2014). How does metacognition contribute to the regulation of learning? An integrative approach. Psihologijske Teme, 23(1), 1–30.
Birney, D. P., Beckmann, J. F., & Wood, R. E. (2012). Precursors to the development of flexible expertise: Metacognitive self-evaluations as antecedents and consequences in adult learning. Learning and Individual Differences, 22(5), 563–574.
