In an eloquent and coherent exposition, Frost makes a convincing case for the argument that reading for meaning is a complex operation and that “orthographic effects in visual word recognition … are the product of the full linguistic environment of the reader (phonology, morphology, and semantic meaning)” (sect. 1, para. 5, emphasis in original). Syntax is not mentioned as part of this environment, yet note that some orthographies may reflect syntactic phenomena. For example, alphabetic systems are sequential, even if non-linear, and adhere to sentential word order.
Frost argues that a theory of reading should be universal and therefore contain “the invariant” among orthographies of the world languages. Frost's program postulates that “a good theory of reading should describe and explicate … the cognitive procedures that are implicated in processing printed words in any orthography, focusing on (1) what characterizes human writing systems, and (2) what characterizes the human cognitive system that processes them. I [i.e., Frost] will label these common cognitive procedures reading universals” (sect. 2.1, para. 1, emphases in original). Yet, what might be universal and invariant among systems that serve speakers of Chinese, English, Hebrew, Japanese, or Finnish?
The search for “reading universals” brings to mind linguistic theories of generative persuasion in which the notion of Universal Grammar (UG) refers to the computational properties of language – any human language (e.g., Chomsky Reference Chomsky2006). But writing systems need not encode UG, since these principles are presumed to be at the core of all human languages and are given “for free” to speakers. It is primarily the language-specific properties, which by definition are not universal, that the system must be attuned to. So it is the
variant
among human languages that orthographies encode, as Frost's examples amply demonstrate.
Frost correctly states that writing systems rely on cognitive parameters. Put differently, writing systems select from among information-encoding principles those that will highlight the relevant features of the language that the message is cast in, and can guarantee fast-enough decoding. Chief among these cognitive devices is order (e.g., the invariance of clusters, seen in the phonological composition of English words), but there are others: For instance, differential prominence among constituent units (e.g., the prominence of Hebrew root consonants vis-à-vis word patterns), or simultaneous processing of parallel systems of codes (e.g., Hebrew letters and diacritics).
Although the foregoing examples, all drawn from Frost, indeed relate to the encoding of information within writing systems, there is nothing linguistic or orthographic about the principles underlying these procedures. Rather, they are general-purpose devices that are available to writing systems but not specific to them. Writing systems exploit such parameters, as do other cognitive systems: for example, navigation in different environments, or conceptual categorization. Thus, rather than setting the stage for “reading universals,” writing systems realize general cognitive principles for encoding information.
Exhibiting one or several of these principles, the system may still not be perfect, for example, in non-optimal circumstances, a writing system may fail to transmit a sufficient amount of information to the reader. Within an interactive, task-oriented cognition, this shortcoming is manageable and can often be resolved, for example, through stronger dependence on context, whether linguistic or even nonlinguistic.
Hebrew provides a case in point. Many Hebrew words have more than one reading when presented without diacritics. Yet, while diacritics enrich the informational value for readers, they slow down writing and require additional expertise. Hebrew readers routinely give up diacritics, relying instead on syntax and lexicon, which generally resolve such ambiguities. Faced with written texts that have no diacritics, readers who do not know Hebrew grammar may not be able to fully decode “unpointed” script. Occasionally, nonlinguistic context can provide support, as well. For example, single, unpointed words used in street signs can be disambiguated within this extralinguistic context.
Thus, optimality need not be achieved within the writing system, as argued by Frost. Rather, it is a product of the entire communicative act. Impoverished systems, for example, unpointed Hebrew, need fallback strategies, such as reliance on context. On the other hand, the non-overlapping word-internal order of letters and full spelling that is characteristic of English orthography, result in a redundant writing system that is well suited for non-optimal reading conditions (e.g., messy scribble, insufficient light, or reduced texts such as in the “Cambridge effect”). Such redundancy frees the reader from strong reliance on sentential context and helps tip the balance in that it allows attention to focus on the complexities of the phonological representations that can be effortful for readers of English.
In an eloquent and coherent exposition, Frost makes a convincing case for the argument that reading for meaning is a complex operation and that “orthographic effects in visual word recognition … are the product of the full linguistic environment of the reader (phonology, morphology, and semantic meaning)” (sect. 1, para. 5, emphasis in original). Syntax is not mentioned as part of this environment, yet note that some orthographies may reflect syntactic phenomena. For example, alphabetic systems are sequential, even if non-linear, and adhere to sentential word order.
Frost argues that a theory of reading should be universal and therefore contain “the invariant” among orthographies of the world languages. Frost's program postulates that “a good theory of reading should describe and explicate … the cognitive procedures that are implicated in processing printed words in any orthography, focusing on (1) what characterizes human writing systems, and (2) what characterizes the human cognitive system that processes them. I [i.e., Frost] will label these common cognitive procedures reading universals” (sect. 2.1, para. 1, emphases in original). Yet, what might be universal and invariant among systems that serve speakers of Chinese, English, Hebrew, Japanese, or Finnish?
The search for “reading universals” brings to mind linguistic theories of generative persuasion in which the notion of Universal Grammar (UG) refers to the computational properties of language – any human language (e.g., Chomsky Reference Chomsky2006). But writing systems need not encode UG, since these principles are presumed to be at the core of all human languages and are given “for free” to speakers. It is primarily the language-specific properties, which by definition are not universal, that the system must be attuned to. So it is the variant among human languages that orthographies encode, as Frost's examples amply demonstrate.
Frost correctly states that writing systems rely on cognitive parameters. Put differently, writing systems select from among information-encoding principles those that will highlight the relevant features of the language that the message is cast in, and can guarantee fast-enough decoding. Chief among these cognitive devices is order (e.g., the invariance of clusters, seen in the phonological composition of English words), but there are others: For instance, differential prominence among constituent units (e.g., the prominence of Hebrew root consonants vis-à-vis word patterns), or simultaneous processing of parallel systems of codes (e.g., Hebrew letters and diacritics).
Although the foregoing examples, all drawn from Frost, indeed relate to the encoding of information within writing systems, there is nothing linguistic or orthographic about the principles underlying these procedures. Rather, they are general-purpose devices that are available to writing systems but not specific to them. Writing systems exploit such parameters, as do other cognitive systems: for example, navigation in different environments, or conceptual categorization. Thus, rather than setting the stage for “reading universals,” writing systems realize general cognitive principles for encoding information.
Exhibiting one or several of these principles, the system may still not be perfect, for example, in non-optimal circumstances, a writing system may fail to transmit a sufficient amount of information to the reader. Within an interactive, task-oriented cognition, this shortcoming is manageable and can often be resolved, for example, through stronger dependence on context, whether linguistic or even nonlinguistic.
Hebrew provides a case in point. Many Hebrew words have more than one reading when presented without diacritics. Yet, while diacritics enrich the informational value for readers, they slow down writing and require additional expertise. Hebrew readers routinely give up diacritics, relying instead on syntax and lexicon, which generally resolve such ambiguities. Faced with written texts that have no diacritics, readers who do not know Hebrew grammar may not be able to fully decode “unpointed” script. Occasionally, nonlinguistic context can provide support, as well. For example, single, unpointed words used in street signs can be disambiguated within this extralinguistic context.
Thus, optimality need not be achieved within the writing system, as argued by Frost. Rather, it is a product of the entire communicative act. Impoverished systems, for example, unpointed Hebrew, need fallback strategies, such as reliance on context. On the other hand, the non-overlapping word-internal order of letters and full spelling that is characteristic of English orthography, result in a redundant writing system that is well suited for non-optimal reading conditions (e.g., messy scribble, insufficient light, or reduced texts such as in the “Cambridge effect”). Such redundancy frees the reader from strong reliance on sentential context and helps tip the balance in that it allows attention to focus on the complexities of the phonological representations that can be effortful for readers of English.