In the last decade, reading research has seen a paradigmatic shift. A new wave of computational models of orthographic processing that offer various forms of noisy position or context-sensitive coding have revolutionized the field of visual word recognition. The influx of such models stems mainly from consistent findings, coming mostly from European languages, regarding an apparent insensitivity of skilled readers to letter order. Underlying the current revolution is the theoretical assumption that the insensitivity of readers to letter order reflects the special way in which the human brain encodes the position of letters in printed words. The present article discusses the theoretical shortcomings and misconceptions of this approach to visual word recognition. A systematic review of data obtained from a variety of languages demonstrates that letter-order insensitivity is neither a general property of the cognitive system nor a property of the brain in encoding letters. Rather, it is a variant and idiosyncratic characteristic of some languages, mostly European, reflecting a strategy of optimizing encoding resources, given the specific structure of words. Since the main goal of reading research is to develop theories that describe the fundamental and invariant phenomena of reading across orthographies, an alternative approach to model visual word recognition is offered. The dimensions of a possible universal model of reading, which outlines the common cognitive operations involved in orthographic processing in all writing systems, are discussed.

Frost has given us good reason to question the universality of existing computational models of reading. Yet, he has not provided arguments showing that all languages share fundamental and invariant reading universals. His goal of outlining the blueprint principles for a universal model of reading is premature. Further, it is questionable whether natural evolution can provide the optimal solutions that Frost invokes.

Frost's claim that universal models of reading require linguistically diverse data is relevant and justified. We support it with evidence demonstrating the extent of the bias towards some Indo-European languages and alphabetic scripts in scientific literature. However, some of his examples are incorrect, and he neglects the complex interaction of writing system and language structure with history and cultural environment.

A universal property of visual word identification is position-invariant letter identification, such that the letter “A” is coded in the same way in CAT and ACT. This should provide a fundamental constraint on theories of word identification, and, indeed, it inspired some of the theories that Frost has criticized. I show how the spatial coding scheme of Colin Davis (2010) can, in principle, account for contrasting transposed letter (TL) priming effects, and at the same time, position-invariant letter identification.

There are universals of language; but is it also true, as the target article claims, that there are universals of reading? We believe there are no such universals, and invite others to refute our claim by providing a list of some universals of reading. If there are no universals of reading, there cannot be a universal model of reading.

I argue, contra Frost, that when prime lexicality and target density are considered, it is not clear that there are fundamental differences between form priming effects in Semitic and European languages. Furthermore, identifying and naming printed words in these languages raises common theoretical problems. Solving these problems and developing a universal model of reading necessitates “cracking” the orthographic input code.

Reading development is integral to a universal model of reading. Developmental research can tell us which factors drive reading acquisition and which are the product of reading. Like adult research, developmental research needs to be contextualised within the language and writing system and it needs to include key cross-linguistic evaluations. This will create a universal model of reading development.

Comparisons across languages have long been a means to investigate universal properties of the cognitive system. Although differences between languages may be salient, it is the underlying similarities that have advanced our understanding of language processing. Frost is not unique in emphasizing that the interaction among linguistic codes reinforces the inadequacy of constructing a model of word recognition where orthographic processes operate in isolation.

The properties of a specific orthography dictate the way people read it. We bring considerations from dyslexia to suggest that the claim can be extended further. First, the effect of orthographic neighborhood density can be extended beyond letter-position encoding and beyond the orthographic lexicon. Second, Hebrew and Arabic differ with respect to letter forms, and hence, in letter-position encoding.

Frost's article advocates for universal models of reading and critiques recent models that concentrate in what has been described as “cracking the orthographic code.” Although the challenge to develop models that can account for word recognition beyond Indo-European languages is welcomed, we argue that reading models should also be constrained by general principles of visual processing and object recognition.

Children's reading and spelling errors show that orthographic learning involves complex interactions with phonology, morphology, and meaning throughout development. Even young children seek to make their visual word recognition strategies linguistically coherent. Orthographic knowledge gained through spelling affects reading, and vice versa. Developmental data support Frost's claim that letter-coding flexibility reflects the optimization of encoding resources in a highly developed system.

Frost proposes a new agenda for reading research, whereby cross-linguistic experiments would uncover linguistic universals to be integrated within a universal theory of reading. We reveal the dangers of following such a call, and demonstrate the superiority of the very approach that Frost condemns.

The principles for universal reading models proposed by Frost correspond to developmental theories, in which neurocognitive constraints and cultural experiences shape development. We question his contention that Hebrew word identification is fundamentally about roots, excluding verbal and nominal word-pattern morphemes; and we propose that readers use all information available in stimuli, adjusting for volume and usefulness.

We agree with Frost that flexible letter-position coding is unlikely to be a universal property of word recognition across different orthographies. We argue that it is particularly unlikely in morphologically rich languages like Finnish. We also argue that dual-route models are not overly flexible and that they are well equipped to adapt to the linguistic environment at hand.

We suggest that the linguistic characteristics of letters also need to be considered to fully understand how a reader processes printed words. For example, studies in Korean showed that unambiguity in the assignment of letters to their appropriate onset, vowel, or coda slot is one of the main sources of the letter-transposition effect. Indeed, the cognitive system that processes Korean is tuned to the structure of the Korean writing system.

Parafoveal preview benefit (PB) is an implicit measure of lexical activation in reading. PB has been demonstrated for orthographic and phonological but not for semantically related information in English. In contrast, semantic PB is obtained in German and Chinese. We propose that these language differences reveal differential resource demands and timing of phonological and semantic decoding in different orthographic systems.

Writing systems encode linguistic information in diverse ways, relying on cognitive procedures that are likely to be general purpose rather than specific to reading. Optimality in reading for meaning is achieved via the entire communicative act, involving, when the need arises, syntax, nonlinguistic context, and selective attention.

In this commentary we concur with Frost's view of the centrality of universal principles in models of word identification. However, we argue that other processes in sentence comprehension also fundamentally constrain the nature of written word identification. Furthermore, these processes appear to be universal. We, therefore, argue that universality in word identification should not be considered in isolation, but instead in the context of other linguistic processes that occur during normal reading.

Frost presents an explanatory theory of reading that generalizes across several languages, based on a revised role of orthographic coding. Perceptual and psychophysical evidence indicates a decay of letter position encoding as a function of the eccentricity of letters (crowding); this factor may account for some of the differences in the languages considered by Frost.

Examples from Chinese, Thai, and Finnish illustrate why researchers cannot always be confident about the precise nature of the word unit. Understanding ambiguities regarding where a word begins and ends, and how to model word recognition when many derivations of a word are possible, is essential for universal theories of reading applied to both developing and expert readers.