Psycholinguistic Approaches to Morphology: Theoretical Issues
Psycholinguistic Approaches to Morphology: Theoretical Issues
- Christina L. GagnéChristina L. GagnéDepartment of Psychology, University of Alberta
Psycholinguistics is the study of how language is acquired, represented, and used by the human mind; it draws on knowledge about both language and cognitive processes. A central topic of debate in psycholinguistics concerns the balance between storage and processing. This debate is especially evident in research concerning morphology, which is the study of word structure, and several theoretical issues have arisen concerning the question of how (or whether) morphology is represented and what function morphology serves in the processing of complex words. Five theoretical approaches have emerged that differ substantially in the emphasis placed on the role of morphemic representations during the processing of morphologically complex words. The first approach minimizes processing by positing that all words, even morphologically complex ones, are stored and recognized as whole units, without the use of morphemic representations. The second approach posits that words are represented and processed in terms of morphemic units. The third approach is a mixture of the first two approaches and posits that a whole-access route and decomposition route operate in parallel. A fourth approach posits that both whole word representations and morphemic representations are used, and that these two types of information interact. A fifth approach proposes that morphology is not explicitly represented, but rather, emerges from the co-activation of orthographic/phonological representations and semantic representations. These competing approaches have been evaluated using a wide variety of empirical methods examining, for example, morphological priming, the role of constituent and word frequency, and the role of morphemic position. For the most part, the evidence points to the involvement of morphological representations during the processing of complex words. However, the specific way in which these representations are used is not yet fully known.
The study of language has played a central role in the study of human cognition. Indeed, Miller (1990, p. 7) made the strong claim that “no general theory of psychology will be adequate if it does not take account of language.” Psycholinguistic research takes into account both the types of knowledge that people have about language and cognitive mechanisms. Morphological processing has been examined in a broad range of areas within psycholinguistics, including word formation, speech production, and acquisition of language. Because the literature spanning all these morphologically relevant subareas of psycholinguistics is so vast, the current article focuses primarily on issues relevant to the mental lexicon. Research on the mental lexicon aims to understand the nature and use of mental representations of words, including information about meaning, pronunciation, orthography, and syntactic characteristics.
Traditionally, access units are representations that mediate between the input stimulus (e.g., the visual information or the auditory information that is presented) and the lexical representations. Access units are more closely associated with the surface form (orthographic and phonological properties) of the word, whereas the lexical representations are more abstract and contain information such as grammatical class. Research on the mental lexicon is highly active, and there has been much debate about the role (if any) of sub-word units in the processing of morphologically complex words such as runner and teacup.
2. Theoretical Issues
A number of long-standing theoretical issues have arisen concerning how (or whether) morphology is represented and what function morphology serves in the processing of complex words. These issues have served as a basis for the various theoretical approaches that have been taken within the field (Section 3) and for guiding psycholinguistic research that seeks to test the various competing theories and assumptions (Section 4).
2.1 Is Morphology Directly Represented in the Language System?
One issue of debate is whether morphology is explicitly represented as a separate level of representation or whether it emerges from semantic and form representations. This issue of storage versus computation has a long history within psycholinguistics and is closely related to debates concerning the production of verbs. Some researchers argue that production is best explained by a system that explicitly uses rules for generating regular forms and direct storage for irregular forms (e.g, Clahsen,1999; Pinker, 1999). Others argue in favor of a neural network in which knowledge about generating verb forms is distributed across the network via associations between phonological features and output nodes, rather than as explicit rules (e.g., Rumelhart & McClelland, 1986).
Although this debate originally centered around the production of past tense forms, the question of storage versus computation has expanded to other aspects of morphology and has been discussed in the context of inflected words, derived words, and more recently, compound words. Inflectional morphemes do not significantly alter the root morpheme (e.g., -s indicates plurality but otherwise does not change the meaning of a root such as boat or house). In contrast, derivational morphemes alter the root morpheme and, for example, can affect the syntactic category of the word (-er changes the root morpheme learn from a verb to a noun). In compound words (e.g., boathouse), both morphemes can function as independent words and the grammatical category of the compound usually corresponds to the head morpheme (in English, the right-most constituent).
A range of theoretical approaches has arisen to account for morphology. Some researchers argue in favor of explicitly including information about a word’s morphological structure in the mental lexicon. For example, Sandra (1990); see also Taft & Kougious (2004) argues that the morpheme is the basic unit of representation in the lexical system and that words are accessed via their morphological constituents. Thus, at the lexical representation level, several words might share a common morphological form (e.g., walk, walking, walked, and walker all share the invariant form walk). Roelofs and Baayen (2002) also argue that morphological representations should be included in psycholinguistic theories and suggest that morphemes might act as planning units during the production of complex words (see also Aronoff, 1994). In these approaches, morphemes are explicitly represented and are either linked to lexical units (e.g., the morphemes walk and ed are linked to walked) or are used to construct the whole-word form via production rules.
In contrast, other researchers argue that morphology has no real existence and instead arises out of systematic relations between form and meaning (e.g., Gonnerman, Seidenberg, & Andersen, 2007; Plaut & Gonnerman, 2000). Depending on the particular theory and also on whether the theory was intended to account for written or spoken language, form could refer to orthographic form or phonological form. Taft and Kougious (2004) suggest that the correlation between form and meaning might be based on a systematically structured orthographic form, such that the initial consonant-vowel-consonant unit of a word (excluding prefixes) becomes associated with a word’s meaning. By this view, morphology is represented in terms of orthographically related clusters of words, rather than as separate abstract morphological representations. Others, such as Bybee (1995) suggest that morphology is best represented in terms of schemas that determine links between whole-word representations of morphologically related words. Schemas are generalizations based on sets of words that share similar patterns of connections between phonological and semantic features. To illustrate, the lexical representations for walk, walking, walks, walked might be linked to each other because they are all members of a set within a common schema. Finally, there are connectionist models (e.g., Plaut & Gonnerman, 2000) that do not assume a symbolic representation and instead posit that the lexical representation of a word is distributed across a set of representation units. These models also claim that morphology arises from correlations between formal (i.e., orthographic or phonological information) and semantic representations.
2.2 At Which Point in Processing Do Morphemes Become Available During Morphological Processing?
In approaches that assume that morphology is explicitly represented within the lexical system, there is much debate concerning the question of whether the decomposition of a complex word into constituent morphemes occurs before or after the access of the word (e.g., Diependaele, Sandra, & Grainger, 2009; Longtin & Meunier, 2005). Theories also differ in whether they assume that morphological decomposition occurs for all word types (e.g., inflected, derived, and compound words), and also in whether decomposition is attempted only for semantically transparent words whose meaning can be derived from the meaning of the constituent morphemes (Giraudo & Grainger, 2001). Finally, there is disagreement about the basis for decomposition, with some researchers positing that decomposition is based only on morpho-orthographic information (e.g., on the surface form of a word) and others positing that semantic information is also used (Beyersmann, Castles, & Coltheart, 2011; Rastle, Davis, & New, 2004).
2.3 How Do Morphemes Contribute to the Activation of a Complex Word?
If morphemes are available prior to the access of a complex word, it still remains to be seen what role they play in the processing of the word and, in particular, in deriving the meaning of that word. Although different claims have been made about the role of constituent representations in activating the lexical and semantic information about the whole word, one thing that seems clear is that if morphemic representations are accessed, they must somehow be integrated during the processing of a complex word.
The most widely used assumption asserts that morphemes are involved in conjunctive activation, whereby the lexical representations of the morphological constituents are activated and assist in activating the lexical representation of the whole word due to facilitatory links between the morphemes and the complex word. For example, the lexical representations of blue and berry are activated and aid in the activation of blueberry, as activation flows from the constituent representations to the representation of the whole word (e.g., Libben, 1998; Sandra, 1990; Schreuder & Baayen, 1995). With respect to semantic representations, this approach assumes that the representations of semantically transparent constituents are activated during processing and help activate the semantic representation of the compound. Thus, the semantic representation of blue aids the retrieval of blueberry, because activation of the lexical representation of blue activates the semantic representation of blue, which in turn activates the semantic representation of blueberry. However, the semantic representations of opaque constituents are assumed to be unconnected to the compound or to have inhibitory links, and thus the semantic representation of straw (which is semantically opaque and does not contribute to the meaning of strawberry) does not aid the retrieval of strawberry.
Alternatively, integration of the constituents might involve semantic composition (Fiorentino & Poeppel, 2007; Gagné & Spalding, 2009). By this view, the language system attempts to construct the meaning of a word based on the semantic representations of the constituents. To illustrate, Schreuder and Baayen (1995) suggest that the meaning of the Dutch word boeken (books) might be created through the union of boek (book) and -en (the plural affix). Likewise, the meaning of cheekbone might be constructed through a conceptual combination process that produces a combined concept based on the individual concepts cheek and bone (Gagné & Spalding, 2004, 2009). Indeed, the processing of compound words appear to be influenced by some of the same factors as novel modifier-noun phrases (e.g., plastic bone), such as the ease with which the two constituents can be linked via a semantic relation (e.g., Gagné & Spalding, 2014a, 2015; Spalding & Gagné, 2014; for a review of this evidence, see Gagné & Spalding, 2014b).
3. Theories: Psycholinguistic Approaches to Morpheme Representation
A wide range of theoretical approaches have arisen concerning the representation and processing of morphologically complex words in terms of the degree to which morphological structure plays a role in the processing of complex words. These approaches have taken different positions concerning the issues described in Section 2. A guiding principle behind many of the architectures proposed for the processing of complex words has been the balance of lexical storage versus morphological computation, and several papers in the literature on complex word processing have been explicitly concerned with establishing how the balance of storage and computation affects lexical processing (e.g., Bertram, Laine, Baayen, Schreuder, & Hyönä, 2000; Kuperman, Bertram, & Baayen, 2010; Kuperman, Schreuder, Bertram, & Baayen, 2009; Libben, 2005).
3.1 Access via Whole-Word Representations
The first approach minimizes processing demands on the system by positing that each word, regardless of whether it is morphologically simple or complex, has its own lexical representation in the mental lexicon and is accessed via this whole-word representation without the involvement of morphemic representations (Butterworth, 1983; Janssen, Bi, & Caramazza, 2008; Lukatela, Carello, & Turvey, 1987; Manelis & Tharp, 1977; Rueckl, Mikolinski, Raveh, Miner, & Mars, 1997). Consequently, simple words and morphologically complex words are both processed using the same set of processes. When direct access to the whole-word representation fails, such as when a word is unfamiliar to the reader or listener, or when a word is novel, then “fall-back” procedures can be evoked to determine the meaning of the word. Butterworth (1983) assumed that these procedures could be rule-based. Others (see Cutler, 1980) suggested that these procedures could be meta-rules, such as looking for words that are similar. Nonetheless, by this approach, direct retrieval of the whole-word form is the default and primary method of lexical access.
According to the whole-word approach, morphological information arises from the semantic information of a word’s lexical representation in that morphologically related words share semantic features. For example, the meaning of rained would be similar to the meaning of rain plus the meaning of the past tense marker -ed. Consequently, morphological analysis arises only at a later stage of processing, after the lexical representation of a word has been retrieved. A variant of this approach allows morphology to be represented in terms of the organization of the lexicon such that words (e.g., walked, walking, walks) that have the same root morpheme (e.g., walk) are connected to each other (Colé, Beauvillain, & Segui, 1989; Drews & Zwitserlood, 1995; Grainger, Colé, & Segui, 1991). This organization, which occurs at the lexical and/or semantic level of representation, allows morphological information to be distinct from orthographic and phonological similarity.
3.2 Access via Composed Forms
The second approach, often known as a full-parsing approach, posits that complex words are represented in the mental lexicon via their constituent morphemes rather than as whole-word representations (Chialant & Caramazza, 1995; Dell, 1986; Frauenfelder & Schreuder, 1991; Laudanna & Burani, 1985; Taft & Forster, 1975, 1976). This general approach has several variants. The theories vary in terms of how decomposition occurs. Some theories posit that the segmentation of words into their morphological components is achieved by the application of parsing rules (e.g., the automatic segmentation of affixes, Taft & Forster, 1975), whereas other theories posit that segmentation is more complex and involves different processing levels (e.g., Taft, 1994).
In addition, theories that rely on morphological decomposition vary in terms of when morphemic information becomes available. Sublexical theories (e.g., Fiorentino & Poeppel, 2007; Longtin & Meunier, 2005; Rastle & Davis, 2008; Rastle, Davis, & New, 2004; Taft & Forster, 1976; Taft & Kougious, 2004) propose that decomposition occurs at an early stage in processing and that morphemic representations are used to access the representation of the whole word. Word recognition begins with obligatory morphological decomposition; for example, the word banker is decomposed into bank and er, and the compound bankroll is decomposed into bank and roll. These constituents are used to derive the meaning of a word through recombination and computation of meaning based on the morphological constituents, or if the meaning cannot be computed from the constituents, by using the constituents to access the whole-word representation. If a word cannot be recognized via its constituents then the system attempts to recognize a word via its full form representation. By this view, morphemic representations are closely connected to the surface form of a word (i.e., to a word’s orthography and phonology).
In contrast, supralexical theories (Diependaele, Sandra, & Grainger, 2009; Giraudo & Grainger, 2001; Grainger, Colé, & Segui, 1991) propose that decomposition occurs at a later processing stage; the full form of a morphologically complex word is activated first, and only later do the constituent representations become available. Thus, morphemic information can influence processing only after the whole-word representation has been accessed. By this view, morphemic representations are more abstract and correspond to base-lexemes rather than to the surface forms of words. By the supralexical view, morphemes are used to capture the correspondence of form and meaning within sets of morphologically related words, but they are not used as a basis for access of the word’s lexical representation. In this respect, this view is similar to the whole-word approach (Section 3.1).
Finally, hybrid models (e.g, Diependaele, Sandra, & Grainger, 2005) posit that both sublexical activation and supralexical activation of morphemic representation occurs. By this view, there are two distinct systems: a morpho-orthographic system supports sublexical activation in the early processing stages and a morpho-semantic system supports supralexical activation in later stages.
3.3 Access via Dual-Route Parallel Processing
The third approach posits that a whole-access route and a decomposition route operate in parallel (e.g., Burani, Salmaso, & Caramazza, 1984; Caramazza, Laudanna, & Romani, 1988; Diependaele, Sandra, & Grainger, 2009; Schreuder & Baayen, 1995; Stanners, Neiser, Hernon, & Hall, 1979). These two access routes operate independently of each other, and processing speed is determined by which route finishes first. Thus, the whole word and constituent representations can be simultaneously available. However, due to the assumption of independence, activation of the whole-word representation does not influence the activation of the constituent morphemes and vice-versa. For example, the Augmented Addressed Morphology Model (Caramazza, Laudanna, & Romani, 1988) assumes that there are two levels of access units: whole-word and decomposed forms. The lexical forms of regularly inflected words and other transparent words are represented as stems (i.e., without affixes). The processing of irregular forms (e.g., went) is guided by orthographic surface form. Another assumption is that known words are processed though whole-word access units, whereas unfamiliar regular words and novel words are processed via morpheme-based access units. Finally, some versions of this approach predict that different word types are represented and processed differently; for example, regular inflected forms might be processed in terms of their root (e.g., walk in the case of the word walking), but derived forms might be processed in terms of their whole-word form.
3.4 Access via Interactive Multiple-Routes Processing
A fourth approach, a multiple-route approach, posits that whole word representations and morphemic representations both play a role in word processing (e.g., Kuperman, Bertram, & Baayen, 2010; Kuperman, Schreuder, Bertram, & Baayen, 2009; see also Libben, 2005). Unlike parallel dual-route models, which do not allow one route to influence the processing of the other route, this approach allows for the processing occurring for the two routes to influence each other. Information is processed in a cascading fashion (i.e., processing at one level does not have to be completed before information is passed to subsequent levels) rather than in a series of strict sequential stages. Consequently, orthographic and semantic processing (for example) could occur concurrently.
For example, Kuperman et al. (2009) proposed that the visual information from a stimulus activates lexical information about that stimulus in memory and that the ease of accessing a word is affected by the amount of information carried by a word (which is based on knowledge about the word itself and of other words with similar morphological paradigms). By this view, morphological structure is a combination of sources of information from different levels (such as characters, direct morphological constituents, abstract morphological representations, and whole word). These different sources of morphological information are not independent and, therefore, can moderate the influence of each other.
Another variant of this type of approach is the form-with-meaning account of Feldman, Milin, Cho, Moscoso del Prado Martín, and O’Connor (2015), in which both form and meaning aspects of a word are simultaneously active and the processing of form need not be completed prior to the processing of meaning (see also Plaut & Gonnerman, 2000; Pulvermüller, 2002). Indeed, by this variant, there is an interaction between meaning and form at early stages of processing. This proposal is consistent with neurophysiological work by Pulvermüller (2002) who found that cortical subnetworks that encode semantics are activated when cortical subnetworks that encode form are activated; these two networks appear to function together, which suggests that there might be concurrent access to both form and semantic information.
3.5 Morphology as the Co-Activation of Form and Meaning in a Distributed Network
A fifth approach proposes that morphology is not explicitly represented, but rather, emerges from the co-activation of orthographic/phonological representations and semantic representations. This approach is similar to the whole-word representation approach (Section 3.1) in that morphology is an emergent property of other types of information, such as orthographic and semantic information; there are direct mappings between the co-activation of formal (orthographic/phonological) and semantic units, and morphology is reflected in the correlations between these units (e.g., Rueckl et al., 1997). However, unlike the whole-word representation approach, which uses symbolic representations (e.g., there is one representation unit for each word), this approach uses distributed representations in which a word is represented by a pattern of activation across a set of nodes within a network.
For example, connectionist models of word processing such as those proposed by Plaut and Gonnerman (2000) and Seidenberg and Gonnerman (2000) do not make use of abstract morphemes or morphological analysis. In these models, lexical knowledge is distributed across simple neuron-like processing units that encode information about the sound and meanings of words. Morphological effects emerge as the connectionist model learns mappings between surface form and meanings. Through these mappings, the system becomes sensitive to correlations between form and meaning.
Topological models of the mental lexicon (e.g., Ferro, Pezzulo, & Pirrelli, 2010) are another example of an approach that uses distributed representation of lexical knowledge and does not explicitly incorporate morphological representations. The mental lexicon is represented as a temporal self-organizing map (TSOM; see also Kohonen, 2001), which consists of a grid of interconnected memory nodes. During training, the TSOM develops a pattern of activation in response to the input words. In this model, lexical storage consists of the activation patterns across the map that become integrated as circuits. For example, if many words contain a particular sequence of letters, such as ing, then a circuit may emerge corresponding to that sequence. Consequently, the processing of words containing that sequence (e.g., working, writing, and running) will tend to recruit and activate the same set of nodes (or circuit). Although morphology is not explicitly represented as a separate level in this type of model, effects of graded morphological structure emerge from the organization of the TSOM.
Another example of the type of approach in which morphemes are not explicitly represented are naive discriminative learning models that have been applied to predicting, for example, inflectional paradigmatic effects (Baayen, Milin, Đurđević, Hendrix, & Marelli, 2011). These models use an incremental learning process (e.g., Rescola & Wagner, 1972) to determine the weights between a set of discriminative features (e.g., letters and letter pairs) and their outcomes (e.g., predicted paradigm or meaning).
4. Empirical Evidence
The methods used for testing the predictions of the competing theories are based on the assumption that mental processing takes time and that observing how a variable influences ease of processing can provide useful insight into the structure and processing of a particular aspect of language, such as the ability to recognize morphologically complex words. For example, if morphemic structure is involved, then experimental manipulations that affect the availability of morphemes, such as providing exposure to a morpheme before presenting it in a morphologically complex word (e.g., preceding the word walked with walk), should alter the time required to process a multi-morphemic word. One of the most common tasks used to study word processing is a lexical decision task in which participants see letter strings and indicate, by pressing one of two computer keys, whether the string corresponds to a real word or to a nonword. Another commonly used task is naming, in which a word is visual presented and participants say the word or else a picture is visual presented and the participants name the object. Eye-tracking procedures are also used in which eye movements are recorded while participants read words (alone or, more typically, in context) and the time spent on various parts of the stimulus are calculated. Some researchers also use electrophysiological brain recording measures, such as EEG and MEG, to examine neural activity in various parts of the brain while participants perform lexical decision or other tasks.
A challenge that arises in isolating morphological effects (see Feldman, 2000, and Rastle & Davis, 2008, for discussions of this problem) is that words that share morphemes also tend to overlap in terms of phonology, orthography, and semantics. Nonetheless, it has been possible to see evidence of the role of morphology that is distinct from form and semantic effects (see, e.g., Assink & Sandra, 2003; Bentin & Feldman, 1990; Frost, Kugler, Deutsch, & Forster, 2005; Gumnior, Bölte, & Zwitserlood, 2006; Rastle & Davis, 2008; Zwitserlood, Bölte, & Dohmes, 2002). Sections 4.1–4.3 contain a discussion of findings concerning three key theoretical questions/issues relevant for assessing the role of morphology.
4.1 Examining Whole-Word Frequency and Constituent Frequency to Evaluate the Time at Which Morphemes Become Available
Frequency refers to how often a particular unit of analysis, such as a word or morpheme, occurs in a given language and is typically measured by counting the number of times that the unit occurs within a particular corpus (e.g.,the SUBTLEX corpus, Brysbaert & New, 2009, or the British National Corpus, 2007). The question of whether frequency influences ease of processing provides insight into the time at which morphological information plays a role. If the influence of whole word frequency comes before and is independent of the influence of morpheme frequency, then this would suggest that the morphological information is used after the lexical representation of the word is accessed.
Numerous studies have found that high frequency words are processed more easily than are low frequency words (Gernsbacher, 1984; Rubenstein & Pollack, 1963; Scarborough, Cortese, & Scarborough, 1977), which suggests that greater exposure to a word either makes that word’s representation more highly activated and, thus, easier to access, or it increases the fluency of processing. Similarly, various experiments have shown that the frequency of morphemes also influence ease of processing (Alegre & Gordon, 1999; Bradley, 1980; Burani & Caramazza, 1987; Colé, Beauvillain, & Segui, 1989; Meunier & Segui, 1999; Niswander, Pollatsek, & Rayner, 2000; Taft, 1979).
The influence of morpheme frequency depends on the type of word (i.e., inflected, derived, and compound words). Responses to inflected and derived words with higher frequency stems are faster than are responses to inflected and derived words with lower frequency stems. This influence has been reported for several languages including English (Taft, 1979), French (Beauvillain, 1996), Italian (Burani & Caramazza, 1987; Burani, Salmaso, & Caramazza, 1984), and Finnish (Lehtonen, Cunillera, Rodriguez-Fornells, Hulten, Tuomainen, & Laine, 2007). Furthermore, the influence of stem frequency occurs for both semantically transparent and opaque stems, which suggests that morpheme effects are not dependent on semantic similarity between the stems and the whole word (Holmes & O’Regan, 1992; Schreuder, Burani, & Baayen, 2003). Compound words also are processed more quickly when they have high frequency morphological constituents than when they have low frequency constituents (see for example, Andrews, 1986; Burani, Salmaso, & Caramazza, 1984). However, the effect of the frequency of the constituent depends on the semantic transparency of the compound; Ji, Gagné, and Spalding (2011) found that higher first-constituent frequency was associated with faster lexical decision times for semantically transparent compounds (i.e., compounds such as teacup for which the meaning of the constituents does contribute to the meaning of the compound) but with slower times for semantically opaque compounds (i.e., compounds such as hogwash for which the meaning of the constituents does not contribute to the meaning of the compound). That is, when the first constituent was more readily available, it was more difficult to process an opaque compound.
Whole-word frequency affects whether the influence of morpheme frequency is facilitatory or inhibitory. For high frequency words, high stem frequency (i.e., frequency of the constituent) is associated with slower processing. In contrast, for low frequency words, high stem frequency is associated with faster processing (Kuperman, Bertram, & Baayen, 2010). In addition, constituents with high word frequencies (or larger morphological families) are associated with faster response times for low-frequency compounds, but are associated with slower response times for high-frequency compounds (see Kuperman, 2013, for a summary). In general, it appears that activation of the constituents speeds the processing of low frequency complex words, but interferes with the processing of higher frequency complex words.
Taken together, the findings suggest that morphological information becomes available during the processing of morphologically complex words, and that the overall ease of processing a complex word depends on the relationships among the frequencies of the whole word and the morphological constituents. Low frequency words are more difficult to access (due to their lower exposure to the processing system in the past), and morphemes could be helpful in accessing lower-frequency word forms. Thus, the overall effect of morpheme frequency would be facilitatory. In contrast, during the processing of a high frequency complex word, access to the word form should be relatively quick and easy, and competition might arise from the whole-word form and the morpheme representations. The net effect of this competition is that the influence of the morpheme frequency on complex word processing is inhibitory. A similar kind of explanation might account for some of the effects of semantic transparency, in that activation of the constituents might be problematic for opaque compounds, but not for transparent ones.
4.2 Using Morphological Priming to Determine Whether Morphemes Become Available During Processing
Priming is an experimental method that involves presenting a word (called the prime) prior to a target word. By systematically varying properties of the prime (for example, by varying whether the prime is morphologically related to the target or whether the prime is only orthographically related to the target) and observing whether these manipulations influence the processing of the target word, researchers are able to test hypotheses about what type of information is available during word recognition. There have been several demonstrations in the literature showing that exposure to a morphologically related word facilitates the processing of a morphologically complex word; for example, dark aids the processing of darkness (e.g., Forster, Davis, Schoknecht, & Carter, 1987; Grainger, Colé, & Segui, 1991). The reverse is also true; exposure to a morphologically complex word aids the processing of the stem, e.g., friendly aids friend (Longtin, Segui, & Hallé, 2003; Marslen-Wilson, Tyler, Waksler, & Older, 1994; Rastle, Davis, Marslen-Wilson, & Tyler, 2000). These effects have also been observed with compound words; exposure to a constituent aids the processing of a compound word (Inhoff, Briihl, & Schwartz, 1996; Jarema, Busson, Nikolova, Tsapkini, & Libben, 1999; Libben, Gibson, Yoon, & Sandra, 2003; Lima & Pollatsek, 1983), and exposure to a compound aids the processing of the constituents (e.g., Zwitserlood, 1994).
There are several variables that influence the size of the morphological priming effect. First, morphological priming is affected by the whole-word frequency of the prime, in that high frequency derived word primes are more beneficial in the subsequent processing of a stem than are low frequency primes (Giraudo & Grainger, 2000; Meunier & Segui, 1999). However, there is some evidence to suggest that word type (i.e., inflected vs. derived words) alters the influence of the prime’s frequency; Raveh (2002) found that high frequency inflected and derived words produced different priming effects; when the primes were high frequency words, responses to target words were faster when preceded by a prime that was a derived word than when preceded by a prime that was an inflected word. However, when the prime was a low frequency word, then the inflected and derived primes produced equivalent priming effects.
Second, semantic transparency also plays a role. Some studies have found that semantically related and unrelated primes produce similar benefits (see Kuperman, 2013; Rastle et al., 2004; Rastle & Davis, 2008). However, whether opaque primes facilitate the processing of the target is more controversial in that the results across various experiments have been inconsistent. For example, some studies using a cross-modal priming task (Longtin, Segui, & Hallé, 2003; Marslen-Wilson, Komisarjevsky Tyler, Waksler, & Older, 1994) have found facilitation for morphologically related words that are semantically related (e.g., hunter-hunt), but not for morphologically related words that were semantically opaque (e.g., gingerly-ginger). Other studies that used priming (e.g., Feldman, 2000; Rastle et al., 2004; see also Rastle & Davis, 2008 for a summary of previous studies on this issue) found equivalent priming from primes that were morphologically structured and were semantically related to the stem (e.g., departure-depart), and from primes that had an opaque morphological structure (e.g., department-depart), and from primes that were orthographically related but not morphologically structured (e.g., brothel-broth). Longtin and Meunier (2005; see also McCormick, Rastle, & Davis, 2008) even found facilitation from morphologically structured nonwords (e.g., darkism-dark), and the priming effect was not affected by whether the nonword was semantically interpretable or not. These results suggest that morphological decomposition is not affected by lexicality or by semantic interpretability.
In contrast, other studies have found that semantically opaque primes were less effective than transparent primes. Diependaele et al. (2009) also used cross-modal priming (with visually presented primes and auditorily presented targets) and observed priming effects for both transparent and opaque words. In their study, the priming effect was larger for semantically transparent words than for opaque words. Similarly, Feldman, Soltano, Pastizzo, and Francis (2004) found that responses to complex words (e.g., casualness) were faster following a semantically transparent prime (e.g., casually) than after a semantically opaque prime (e.g., casualty). Morris, Frank, Grainger, and Holcomb (2007) report the same pattern of results using event-related potential (ERP) data. In terms of compound words, the results concerning semantic transparency have been more consistent and less controversial. Numerous studies have found that exposure to a compound influences the processing of a constituent of that compound, regardless of whether the constituent is transparent or opaque (e.g., Zwitserlood, 1994). Exposure to both transparent and opaque constituent also aids the processing of a compound (Monsell, 1985; Sandra, 1990). These results suggest that the constituent representations become available during the processing of compound words and aid in accessing the compound’s lexical presentation. The finding that this priming occurs even for semantically opaque constituents challenges the idea that the influence of morphology can be explained solely by the overlap between form and semantics.
4.3 Examining Whether Morphemic Representations Are Position-Specific
The findings outlined in Sections 4.1 and 4.2 suggest that morphemes are available during the processing of complex words, and this raises questions about whether morpheme processing is position-specific. The degree to which a morpheme is position-specific appears to depend on the type of morpheme. Suffixes, for example, are strongly position-specific such that when a morpheme that is always used as a suffix appears in a different position (e.g., at the beginning of a word), it is not processed as a morpheme. Suffix morphemes (e.g., ful) only influenced processing when they appeared in the correct (i.e., word-final) position; Crepaldi, Rastle, and Davis (2010) found that, in a lexical decision task where participants indicate whether an item is a word, morphologically complex nonwords (e.g., gasful) take longer to reject than orthographic controls (e.g., gasfil). However, when the constituents were reversed (e.g., fulgas vs. filgas), there was no processing difference.
In contrast, free morphemes are much less position-specific. Zwitserlood, Bölte, and Dohmes (2002) found that picture-naming latencies were faster when the distractor words (presented seven or more trials earlier) were morphologically related to the picture and that this facilitation occurred regardless of the position of the shared morpheme; naming a picture of a rose was aided by previous exposure to either rosebud or tearose, which suggests that processing activated the full word corresponding to the stem morpheme (i.e., __rose or rose__ activated the word rose). Another finding that supports this claim is that reversed compounds take longer to reject in lexical decision experiments, compared to matched nonword controls (Crepaldi, Rastle, Davis, & Lupker, 2013; Shoolman & Andrews, 2003; Taft, Zhu, & Peng, 1999). For example, moonhoney, formed by reversing the compound honeymoon, took longer to reject than moonbasin. Even though reversed, the morphemes were able to activate the existing compound (honeymoon), which made it more difficult to correctly respond “nonword” to the reversed compound. In contrast, this effect did not occur for reversed monomorphemic words (Crepaldi et al., 2013); rickmave does not affect the processing of maverick. Both sets of findings suggest that free morphemes become activated regardless of position.
However, other evidence suggests that the processing of free morphemes is affected by position. Duñabeitia, Laka, Perea, and Carreiras (2009) conducted priming experiments in Basque and found facilitation from compounds that shared a constituent in a different position; responses to mendikate (mendi+kate, ‘mountain+chain,’ meaning ‘mountain range’) were faster when preceded by sumendi (su+mendi, ‘fire+mountain,’ meaning ‘volcano’) even though the shared constituent (mendi) was in a different position. However, the facilitation from the different-position items was smaller than the facilitation produced by same-position items (e.g., lanpostu ‘workplace’ as a prime for the target lanordu ‘working hour’), which suggests that the identification of the morphemes is sensitive to position. This result is consistent with Gagné, Spalding, Figueredo, and Mullaly’s (2009) finding that responses to a compound (e.g., fingernail) were faster when preceded by a prime that used a shared constituent in the same position (e.g., finger cymbals) than when preceded by a prime that used a shared constituent in a different position (e.g., ring finger). Further evidence that the processing of free morphemes might be partially position-specific comes from results showing that response times to compounds are only affected by morphological family members in the same position (Gagné & Spalding, 2009). Morphological family members consist of all derived and compound words in the database that include the constituent. To illustrate, response times to doghouse were influenced by family members of the form dog__ and __house (e.g., dogcatcher and boathouse), but not by words of the form __dog and house__ (e.g., bulldog and housework). De Jong, Feldman, Schreuder, Pastizzo, and Baayen (2002) also found an influence of positional family frequency (i.e., the frequency of all family members in which a constituent appears in the same position as the target compound) and positional family size (i.e., the number of members in which a constituent appears in the same position as the target compound) on the processing of Dutch and English compounds. In sum, the processing of free morphemes in compounds is sensitive to the position of the morpheme, though not to the same extent as the processing of bound morphemes such as suffixes.
5. Critical Analysis and Future Directions
Debate about the most appropriate way to account for morphology continues, and there is not yet strong consensus about how best to incorporate the observed effects of morphology into psycholinguistic theories. However, for the most part, psycholinguistic evidence suggests the involvement of morphemes in the processing of complex words. These effects are not solely attributable to orthography. In addition, morphological effects are observed for semantically opaque morphemes, which suggests that morphological effects are not attributable solely to overlap in semantics among morphologically related words.
At first glance, it might seem that the empirical evidence contains several ambiguous results and inconsistencies, in that different studies reveal somewhat different results depending on the nature of the design and of the materials. However, a clearer picture emerges if the assumptions used to interpret the data are reconsidered. Traditionally, the empirical findings have been interpreted using a particular set of assumptions; namely, that the mental lexicon contains a relatively stable set of representations. Changes to the system can occur as new words are added to the system, but the fundamental architecture and the nature of the processing is relatively stable and unchanging. When the data are interpreted using this underlying assumption, then the findings are seen as being indicative of the nature of the representations. Thus, the lack of evidence for a particular factor is interpreted as the absence of a particular type of information, or, in other cases, as a lack of a connection between representations (e.g., between a morphemic representation and a whole word representation).
However, the diverse set of empirical findings might actually be more reflective of a framework that has not yet been fully incorporated into existing theoretical approaches. It has already been shown that the nature of the processing is influenced by the mixture of the experimental stimuli that are used (e.g., Keuleers, Diependaele, & Brysbaert, 2010) and by the task (e.g., Duñabeitia, Kinoshita, Carreiras, & Norris, 2011). More recent evidence (Gagné & Spalding, 2014c) has shown that the lexical system can adapt very rapidly to different processing demands. The seemingly inconsistent or inconclusive results might be indicative of this highly adaptive and dynamic language system. Rather than interpreting empirical effects as reflecting solely the architecture of the system (e.g., in terms of whether certain representations are or are not present, or whether representations are or are not connected to each other), it is worth considering that effects are also reflective of processing (see Gagné & Spalding, 2014c, and Libben, 2005, for fuller discussions of this possibility). In this framework, the cognitive system allocates different attentional processing and engages different sets of processes to meet the demands of the current context.
Of the five main theoretical approaches outlined in Section 3, the multi-route approach (Section 3.4) comes closest to expressing this type of framework in that it does allow for multiple sources of information from different levels of representations (e.g., orthographic, morphemic, semantic, etc.) to be available at the same time and to influence each other. However, the multi-route approach is a fixed architecture and does not dynamically adapt to processing demands. By also considering processing in addition to representational implications of the data, new insights into the language system might be discovered that would be able to unite disparate sets of effects. Re-examining assumptions about representation and processing might lead to fruitful avenues of research as psycholinguists seek to interpret and account for the vast amount of findings that have been generated. Also, re-examination of these assumptions might aid in the development of alternative theoretical frameworks that are flexible enough to accommodate the variety of findings concerning the nature of morphology.
- Assink, E. M. H., & Sandra, D. (2003). Reading complex words: Cross-language studies. New York: Kluwer Academic.
- Bauer, L., Lieber, R., & Plag, I. (2013). The Oxford reference guide to English morphology. Oxford: Oxford University Press.
- Gaskell, G. (2007). Oxford handbook of psycholinguistics. Oxford: Oxford University Press.
- Jarema, G., & Libben, L. (2007). The mental lexicon: Core perspectives. Amsterdam: Elsevier.
- Traxler, M., & Gernsbacher, M. A. (2011). Handbook of psycholinguistics. San Diego, CA: Academic Press.
- Alegre, M., & Gordon, P. (1999). Frequency effects and the representational status of regular inflections. Journal of Memory and Language, 40, 41–61.
- Andrews, S. (1986). Morphological influences on lexical access: Lexical or nonlexical effects? Journal of Memory and Language, 25(6), 726–740.
- Aronoff, M. (1994). Morphology by itself. Cambridge, MA: MIT Press.
- Assink, E. M. H., & Sandra, D. (2003). Reading complex words: Cross-language studies. New York: Kluwer Academic.
- Baayen, R. H., Milin, P., Đurđević, D. F., Hendrix, P., & Marelli, M. (2011). An amorphous model for morphological processing in visual comprehension based on naive discriminative learning. Psychological Review, 118(3), 438–481.
- Beauvillain, C. (1996). The integration of morphological and whole-word form information during eye fixations on prefixed and suffixed words. Journal of Memory and Language, 35(6), 801–820.
- Bentin, S., & Feldman, L. B. (1990). The contribution of morphological and semantic relatedness to repetition priming at short and long lags: Evidence from Hebrew. Quarterly Journal of Experimental Psychology: Human Experimental Psychology, 42(4), 693–711.
- Bertram, R., Laine, M., Harald Baayen, R., Schreuder, R., & Hyönä, J. (2000). Affixal homonymy triggers full-form storage, even with inflected words, even in a morphologically rich language. Cognition, 74(2), B13–B25.
- Beyersmann, E., Castles, A., & Coltheart, M. (2011). Early morphological decomposition during visual word recognition: Evidence from masked transposed-letter priming. Psychonomic Bulletin & Review, 18, 937–942.
- Bradley, D. (1980). Lexical representation of derivational relation. In M. Aronoff & M. L. Kean (Eds.), Juncture (pp. 37–55). Saratoga, CA: Anma Libri.
- The British National Corpus, version 3 (BNC XML Edition). (2007). Distributed by University of Oxford Text Archive on behalf of the BNC Consortium.
- Brysbaert, M., & New, B. (2009). Moving beyond Kucera and Francis: A critical evaluation of current word frequency norms and the introduction of a new and improved word frequency measure for American English. Behavior Research Methods, 41(4), 977–990.
- Burani, C., & Caramazza, A. (1987). Representation and processing of derived words. Language and Cognitive Processes, 2(3–4), 217–227.
- Burani, C., Salmaso, D., & Caramazza, A. (1984). Morphological structure and lexical access. Visible Language, 18, 342–352.
- Butterworth, B. (1983). Lexical representation. In B. Butterworth (Ed.), Language production (Vol. 2, pp. 257–294). San Diego, CA: Academic Press.
- Bybee, J. (1995). Regular morphology and the lexicon. Language and Cognitive Processes, 10(5), 425–455.
- Caramazza, A., Laudanna, A., & Romani, C. (1988). Lexical access and inflectional morphology. Cognition, 28(3), 297–332.
- Chialant, D., & Caramazza, A. (1995). Where is morphology and how is it processed? The case of written word recognition. In L. B. Feldman (Ed.), Morphological aspects of language processing (pp. 55–76). Hillsdale, NJ: Lawrence Erlbaum.
- Clahsen, H. (1999). Lexical entries and rules of language: A multi-disciplinary study of German inflection. Behavioral and Brain Sciences, 22, 991–1060.
- Colé, P., Beauvillain, C., & Segui, J. (1989). On the representation and processing of prefixed and suffixed derived words: A differential frequency effect. Journal of Memory and Language, 28(1), 1–13.
- Crepaldi, D., Rastle, K., & Davis, C. J. (2010). Morphemes in their place: Evidence for position specific identification of suffixes. Memory & Cognition, 38(3), 312–321.
- Crepaldi, D., Rastle, K., Davis, C. J., & Lupker, S. J. (2013). Seeing stems everywhere: Position-independent identification of stem morphemes. Journal of Experimental Psychology. Human Perception and Performance, 39(2), 510–525.
- Cutler, A. (1980). Productivity in word formation. In J. Kreiman, & A. E. Ojeda (Eds.), Papers from the Sixteenth Regional Meeting, Chicago Linguistic Society (pp. 45–51). Chicago: Chicago Linguistics Society.
- de Jong, N. H., Feldman, L. B., Schreuder, R., Pastizzo, M., & Baayen, R. H. (2002). The processing and representation of Dutch and English compounds: Peripheral morphological and central orthographic effects. Brain and Language, 81(1–3), 555–567.
- Dell, G. S. (1986). A spreading-activation theory of retrieval in sentence production. Psychological Review, 93(3), 283–321.
- Diependaele, K., Sandra, D., & Grainger, J. (2005). Masked cross-modal morphological priming: Unravelling morpho-orthographic and morpho-semantic influences in early word recognition. Language and Cognitive Processes, 20(1–2), 75–114.
- Diependaele, K., Sandra, D., & Grainger, J. (2009). Semantic transparency and masked morphological priming: The case of prefixed words. Memory & Cognition, 37(6), 895–908.
- Drews, E., & Zwitserlood, P. (1995). Morphological and orthographic similarity in visual word recognition. Journal of Experimental Psychology. Human Perception, and Performance, 21(5), 1098–1116.
- Duñabeitia, J. A., Kinoshita, S., Carreiras, M., & Norris, D. (2011). Is morpho-orthographic decomposition purely orthographic? Evidence from masked priming in the same–different task. Language and Cognitive Processes, 26(4–6), 509–529.
- Duñabeitia, J. A., Laka, I., Perea, M., & Carreiras, M. (2009). Is milkman a superhero like batman? Constituent morphological priming in compound words. European Journal of Cognitive Psychology, 21(4), 615–640.
- Feldman, L. B. (2000). Are morphological effects distinguishable from the effects of shared meaning and shared form? Journal of Experimental Psychology. Learning, Memory, and Cognition, 26(6), 1431–1444.
- Feldman, L. B., Milin, P., Cho, K. W., Moscoso Del Prado Martín, F., & O’Connor, P. A. (2015). Must analysis of meaning follow analysis of form? A time course analysis. Frontiers in Human Neuroscience, 9, 111.
- Feldman, L. B., Soltano, E. G., Pastizzo, M. J., & Francis, S. E. (2004). What do graded effects of semantic transparency reveal about morphological processing? Brain and Language, 90(1–3), 17–30.
- Ferro, M., Pezzulo, G., & Pirrelli, V. (2010). Morphology, memory, and the mental lexicon. Lingue e Linguaggio, 9(2), 199–238.
- Fiorentino, R., & Poeppel, D. (2007). Compound words and structure in the lexicon. Language and Cognitive Processes, 22(7), 953–1000.
- Forster, K. I., Davis, C., Schoknecht, C., & Carter, R. (1987). Masked priming with graphemically related forms: Repetition or partial activation? The Quarterly Journal of Experimental Psychology, 39(2), 211–251.
- Frauenfelder, U. H., & Schreuder, R. (1991). Constraining psycholinguistic models of morphological processing and representation: The role of productivity. In G. Booji & J. van Marle (Eds.), Yearbook of morphology (pp. 165–183). Dordrecht, Netherlands: Kluwer Academic.
- Frost, R., Kugler, T., Deutsch, A., & Forster, K. I. (2005). Orthographic structure versus morphological structure: Principles of lexical organization in a given language. Journal of Experimental Psychology. Learning, Memory, and Cognition, 31(6), 1293–1326.
- Gagné, C. L., & Spalding, T. L. (2004). Effect of relation availability on the interpretation and access of familiar noun-noun compounds. Brain and Language, 90, 478–486.
- Gagné, C. L., & Spalding, T. L. (2009). Constituent integration during the processing of compound words: Does it involve the use of relational structures? Journal of Memory and Language, 60, 20–35.
- Gagné, C. L., & Spalding, T. L. (2014a). Relational diversity and ease of processing for opaque and transparent English compounds. In F. Rainer, F. Gardani, H. C. Luschutzky, & W. U. Dressler (Eds.), Morphology and Meaning (pp. 153–162). Amsterdam: John Benjamins.
- Gagné, C. L., & Spalding, T. L. (2014b). Conceptual composition: The role of relational competition in the comprehension of modifier-noun phrases and noun-noun compounds. In B. H. Ross (Ed.), The psychology of learning and motivation (pp. 97–130). New York: Elsevier.
- Gagné, C. L., & Spalding, T. L. (2014c). Adaptation effects in lexical processing. Suvremena Lingvistika (Contemporary Linguistics), 40(78), 127–149.
- Gagné, C. L. & Spalding, T. L. (2015). Noun-noun compounds. P. O. Müller, I. Ohnheiser, S. Olsen, F. Rainer (Eds.), Word-formation: An international handbook of the languages of Europe (pp. 1143–1159). Berlin: De Gruyter Mouton.
- Gagné, C. L., Spalding, T. L., Figueredo, L., & Mullaly, A. C. (2009). Does snow man prime plastic snow? The effect of constituent position in using relational information during the interpretation of modifier-noun phrases. The Mental Lexicon, 4(1), 41–76.
- Gernsbacher, M. A. (1984). Resolving 20 years of inconsistent interactions between lexical familiarity and orthography, concreteness, and polysemy. Journal of Experimental Psychology: General, 113(2), 256–281.
- Giraudo, H., & Grainger, J. (2000). Effects of prime word frequency and cumulative root frequency in masked morphological priming. Language and Cognitive Processes, 15(4–5), 421–444.
- Giraudo, H., & Grainger, J. (2001). Priming complex words: Evidence for supralexical representation of morphology. Psychonomic Bulletin & Review, 8(1), 127–131.
- Gonnerman, L. M., Seidenberg, M. S., & Andersen, E. S. (2007). Graded semantic and phonological similarity effects in priming: Evidence for a distributed connectionist approach to morphology. Journal of Experimental Psychology. General, 136(2), 323–345.
- Grainger, J., Colé, P., & Segui, J. (1991). Masked morphological priming in visual word recognition. Journal of Memory and Language, 30(3), 370–384.
- Gumnior, H., Bölte, J., & Zwitserlood, P. (2006). A chatterbox is a box: Morphology in German word production. Language & Cognitive Processes, 21(7–8), 920–944.
- Holmes, V. M., & O’Regan, J. K. (1992). Reading derivationally affixed French words. Language and Cognitive Processes, 7(2), 163–192.
- Inhoff, A. W., Briihl, D., & Schwartz, J. (1996). Compound word effects differ in reading, on-line naming, and delayed naming tasks. Memory & Cognition, 24(4), 466–476.
- Janssen, N., Bi, Y., & Caramazza, A. (2008). A tale of two frequencies: Determining the speed of lexical access for Mandarin Chinese and English compounds. Language and Cognitive Processes, 23(7), 1191–1223.
- Jarema, G., Busson, C., Nikolova, R., Tsapkini, K., & Libben, G. (1999). Processing compounds: A cross-linguistic study. Brain and Language, 68, 362–369.
- Ji, H., Gagné, C. L., & Spalding, T. L. (2011). Benefits and costs of lexical decomposition and semantic integration during the processing of transparent and opaque English compounds. Journal of Memory and Language, 65, 406–430.
- Keuleers, E., Diependaele, K., & Brysbaert, M. (2010). Practice effects in large-scale visual word recognition studies: A lexical decision study on 14,000 Dutch mono- and disyllabic words and nonwords. Frontiers in Psychology, 1, 174.
- Kohonen, T. (2001). Self-organizing maps. New York: Springer.
- Kuperman, V. (2013). Accentuate the positive: Semantic access in English compounds. Frontiers in Psychology, 4(203), 1–10.
- Kuperman, V., Bertram, R., & Baayen, R. H. (2010). Processing trade-offs in the reading of Dutch derived words. Journal of Memory and Language, 62(2), 83–97.
- Kuperman, V., Schreuder, R., Bertram, R., & Baayen, R. H. (2009). Reading polymorphemic Dutch compounds: Toward a multiple route model of lexical processing. Journal of Experimental Psychology: Human Perception and Performance, 35, 876–895.
- Laudanna, A., & Burani, C. (1985). Address mechanisms to decomposed lexical entries. Linguistics, 23(5), 775–792.
- Lehtonen, M., Cunillera, T., Rodriguez-Fornells, A., Hulten, A., Tuomainen, J., & Laine, M. (2007). Recognition of morphologically complex words in Finnish: Evidence from event-related potentials. Brain Research, 1148, 123–137.
- Libben, G. (1998). Semantic transparency in the processing of compounds: Consequences for representation, processing, and impairment. Brain and Language, 61(1), 30–44.
- Libben, G. (2005). Everything is psycholinguistics: Material and methodological considerations in the study of compound processing. Canadian Journal of Linguistics, 50, 267–283.
- Libben, G., Gibson, M., Yoon, Y. B., & Sandra, S. (2003). Compound fracture: The role of semantic transparency and morphological headedness. Brain and Language, 84, 50–64.
- Lima, S. D., & Pollatsek, A. (1983). Lexical access via an orthographic code? The basic orthographic syllabic structure (BOSS) reconsidered. Journal of Verbal Learning and Verbal Behavior, 22(3), 310–332.
- Longtin, C.-M., & Meunier, F. (2005). Morphological decomposition in early visual word processing. Journal of Memory and Language, 53(1), 26–41.
- Longtin, C. M., Segui, J., & Hallé, P. A. (2003). Morphological priming without morphological relationship. Language and Cognitive Processes, 18(3), 313–334.
- Lukatela, G., Carello, C., & Turvey, M. T. (1987). Lexical representation of regular and irregular inflected nouns. Language and Cognitive Processes, 2(1), 1–17.
- Manelis, L., & Tharp, D. A. (1977). The processing of affixed words. Memory & Cognition, 5(6), 690–695.
- Marslen-Wilson, W., Komisarjevsky Tyler, L., Waksler, R., & Older, L. (1994). Morphology and meaning in the English mental lexicon. Psychological Review, 101(1), 3–33.
- McCormick, S. F., Rastle, K., & Davis, M. H. (2008). Is there a “fete” in “fetish”? Effects of orthographic opacity on morpho-orthographic segmentation in visual word recognition. Journal of Memory and Language, 58(2), 307–326.
- Meunier, F., & Segui, J. (1999). Morphological priming effect: The role of surface frequency. Brain and Language, 68(1–2), 54–60.
- Miller, G. A. (1990). The place of language in a scientific psychology. Psychological Science, 1(1), 7–14.
- Monsell, S. (1985). Repetition and the lexicon. In A. W. Ellis (Ed.), Progress in the psychology of language (Vol. 2, pp. 147–195). Hillsdale, NJ: Lawrence Erlbaum.
- Morris, J., Frank, T., Grainger, J., & Holcomb, P. J. (2007). Semantic transparency and masked morphological priming: An ERP investigation. Psychophysiology, 44(4), 506–521.
- Niswander, E., Pollatsek, A., & Rayner, K. (2000). The processing of derived and inflected suffixed words during reading. Language and Cognitive Processes, 15(4–5), 389–420.
- Pinker, S. (1999). Words and rules. London: Weidenfeld & Nicolson.
- Plaut, D. C., & Gonnerman, L. M. (2000). Are non-semantic morphological effects incompatible with a distributed connectionist approach to lexical processing? Language and Cognitive Processes, 15(4–5), 445–485.
- Pulvermüller, F. (2002). The neuroscience of language: On brain circuits of words and serial order. Cambridge, U.K.: Cambridge University Press.
- Rastle, K., & Davis, M. H. (2008). Morphological decomposition based on the analysis of orthography. Language and Cognitive Processes, 23(7/8), 942–971.
- Rastle, K., Davis, M. H., & New, B. (2004). The broth in my brother’s brothel: Morpho-orthographic segmentation in visual word recognition. Psychonomic Bulletin & Review, 11(6), 1090–1098.
- Rastle, K., Davis, M. H., Marslen-Wilson, W. D., & Tyler, L. K. (2000). Morphological and semantic effects in visual word recognition: A time-course study. Language and Cognitive Processes, 15(4–5), 507–537.
- Raveh, M. (2002). The contribution of frequency and semantic similarity to morphological processing. Brain and Language, 81(1–3), 312–325.
- Rescola, R. A., & Wagner, A. R. (1972). A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and nonreinforcement. In A. H. Black & W. F Prokasy (Eds.), Classical Conditioning II (pp. 64–99), New York: Apple-Century-Crofts.
- Roelofs, A., & Baayen, H. (2002). Morphology by itself in planning the production of spoken words. Psychonomic Bulletin & Review, 9(1), 132–138.
- Rubenstein, H., & Pollack, I. (1963). Word predictability and intelligibility. Journal of Verbal Learning and Verbal Behavior, 2, 147–158.
- Rueckl, J. G., Mikolinski, M., Raveh, M., Miner, C. S., & Mars, F. (1997). Morphological priming, fragment completion, and connectionist networks. Journal of Memory and Language, 36(3), 382–405.
- Rumelhart, D. E., & McClelland, J. L. (1986). On learning the past tense of English verbs. In J. L. McClelland, D. E. Rumelhart, & the PDP Research Group (Eds.), Parallel Distributed Processing: Explorations in the Microstructure of Cognition, Vol. 2: Psychological and Biological Models (pp. 216–271). Cambridge, MA: MIT Press.
- Sandra, D. (1990). On the representation and processing of compound words: Automatic access to constituent morphemes does not occur. The Quarterly Journal of Experimental Psychology, 42A(3), 529–567.
- Scarborough, D. L., Cortese, C., & Scarborough, H. S. (1977). Frequency and repetition effects in lexical memory. Journal of Experimental Psychology. Human Perception, and Performance, 3(1), 1–17.
- Schreuder, R., & Baayen, R. H. (1995). Modeling morphological processing. In L. B. Feldman (Ed.), Morphological aspects of language processing (pp. 131–154). Hillsdale, NJ: Lawrence Erlbaum.
- Schreuder, R., Burani, C., & Baayen, R. H. (2003). Parsing and semantic opacity. In E. M. H. Assink & D. Sandra (Eds.), Reading complex words: Cross-language studies (pp. 159–189). Dordrecht, Netherlands: Kluwer Academic.
- Seidenberg, M. S., & Gonnerman, L. M. (2000). Explaining derivational morphology as the convergence of codes. Trends in Cognitive Sciences, 4(9), 353–361.
- Shoolman, N., & Andrews, S. (2003). Racehorses, reindeer, and sparrows: Using masked priming to investigate morphological influences on compound word identification. In S. Kinoshita & S. J. Lupker (Eds.), Masked priming: The state of the art (pp. 241–278). New York: Psychology Press.
- Spalding, T. L., & Gagné, C. L. (2014). Relational diversity affects ease of processing even for opaque English compounds. The Mental Lexicon, 9(1), 48–66.
- Stanners, R. F., Neiser, J. J., Hernon, W. P., & Hall, R. (1979). Memory representation for morphologically related words. Journal of Verbal Learning and Verbal Behavior, 18(4), 399–412.
- Taft, M. (1979). Lexical access via an orthographic code: The basic orthographic syllabic structure (BOSS). Journal of Verbal Learning and Verbal Behavior, 18(1), 21–39.
- Taft, M. (1994). Interactive-activation as a framework for understanding morphological processing. Language and Cognitive Processes, 9(3), 271–294.
- Taft, M., & Forster, K. (1976). Lexical storage and retrieval of polymorphemic and polysyllabic words. Journal of Verbal Learning and Verbal Behavior, 15, 607–620.
- Taft, M., & Forster, K. I. (1975). Lexical storage and retrieval of prefixed words. Journal of Verbal Learning and Verbal Behavior, 14(6), 638–647.
- Taft, M., & Kougious, P. (2004). The processing of morpheme-like units in monomorphemic words. Brain and Language, 90(1–3), 9–16.
- Taft, M., Zhu, X., & Peng, D. (1999). Positional specificity of radicals in Chinese character recognition. Journal of Memory and Language, 40(4), 498–519.
- Zwitserlood, P. (1994). The role of semantic transparency in the processing and representation of Dutch compounds. Language and Cognitive Processes, 9(3), 341–368.
- Zwitserlood, P., Bölte, J., & Dohmes, P. (2002). Where and how morphologically complex words interplay with naming pictures. Brain and Language, 81, 358–367.