Derivational morphology
How do we process words? We know that words are built out of a root or stem, i.e. a base, and one or more affixes, i.e. infixes, prefixes or suffixes. The field that is concerned with how lexical representations such as words are created by combining different roots and affixes to give rise to polymorphemic words is known as derivational morphology. Polymorphemic means that a word consists of at least two morphemes. A morpheme is a meaningful linguistic unit that cannot be broken down any further into any more, smaller meaningful units. For example, the word ‘polymorphemic’ consists of the three linguistic units that are the prefix ‘poly’, the base ‘morpheme’ and the suffix ‘ic’. This word therefore can be divided into three meaningful linguistic units while units such as ’poly’ or ‘ic’ cannot be broken down into any more meaningful units.
In order to go from the morphological basic units to the derived words, one needs to take a certain number of steps. This number of steps often varies. For example, in the case of the derived form ‘nationality’, it is clear that this word derives from ‘national’, which in turn is derived from the noun ‘nation’. In this case therefore two steps are needed to move between ‘nation’ and ‘nationality’. This can be formally expressed in the following way: nationality <national<nation-N. In addition to two-step forms, there are also one-step forms where only one step is required to go from the basic unit to the derived form. For example, in the case of ‘development’, it is clear that it is derived from ‘develop’, which is a verb. This can be expressed in the following way: development<develop-V. Another example of one-step words is ‘soaking’, which morphologically is derived from the verb ‘soak’: soaking < soak-V. The verb root therefore is the basic form for one-step words.
Similar to ’soaking’, another example that consists of a verb root (e.g. ‘soak’) and the ‘ing’ form, is ‘eating’. In this sense, ‘eating’ and ‘soaking’ have the same surface structure ‘ing’ while having different verb roots. In both cases, only one step is needed to go from the basic form that is the verb root (e.g. ‘eat’, or ‘soak’) to the derived word (e.g. ‘eating’, or ‘soaking’), which makes them both one-step words.
Zero-derivation
When one looks at two-step words such as ‘bridging’, it is clear that at the surface both one-step and two-step forms have the same structure that is the surface structure ‘-ing’. ‘Bridging’ is a two-step word because it can be derived from the verb ‘bridge’, which in turn is derived from the noun ‘bridge’. This can be expressed in the following way: bridging<bridge-V<bridge-N. Here one can see that for two-step words the verb root (bridge-V) is zero derived from a basic noun (bridge-N).
Morphological processing where the derivational steps are not overtly marked as in this case (e.g. bridge-V and bridge-N) is called ‘zero-derivation’ (Aronoff, 1980).
Examples of Zero Derivations
Noun Verb
the bridge to bridge
the knot to knot
the skate to skate
the bike to bike
the work to work
Zero-derivation is a word class alternation because you have semantically related pairs of homophonous forms that differ in parts of speech. For example, there is a difference in how you would use the words ‘a bridge’ or ‘to bridge’ or ‘a knot’ and ‘to knot’ within a sentence. In both cases the words sound the same but both would be used in different contexts. Specifically, ‘bridge’ within a sentence would be used as a noun, while ‘to bridge’ within a sentence would be used as a verb. Similarly, while ‘a knot’ would be used within a sentence as a noun, ‘to knot’ would be used as a verb.
Some theories argue that ‘to knot’ is covertly derived from ‘knot’ or that ‘development’ is covertly derived from ‘develop’. Here it is obvious that in a covert derivational relationship, the derived form is morphologically more complex than the base. On the other hand, you have theories that argue that such pairs are two forms of a single lexeme that has no inherent word class. In an integrated approach, you would have grammar differentiating between distinct forms of zero-related pairs on the basis of their underlying morphological relationships.
Processing morphologically complex derivations
Morphologically complex derivations are decomposed automatically while we process them mentally (Marslen-Wilson et al., 1994). How are morphologically complex derived words processed in our brain? It has been observed that the process of decomposing morphologically complex derivations causes increased activity in our brain (Gold & Rastle, 2007). Specifically, complex derivations cause more brain activity than simple derivations (Pliatsikas et al., 2014). Accordingly, more brain activity was reported for one-step and two-step forms versus simple forms. This was observed within areas of the brain that are implicated in morphological processing such as the left inferior frontal gyrus (LIFG). Similarly, more brain activity for two-step forms was reported than for one-step forms. Specifically, there was more increased brain activity in the LIFG for two-step versus one-step nouns that was accompanied by heightened activity in occipital regions and bilateral superior temporal regions.
What tasks have been used in research to investigate morphological processing? One approach is to use a single-word presentation paradigm, in which participants are given a lexical decision task. Here derived forms of words are presented, and participants decide via a button press whether the word they see on the screen is a word or not. This task allows you to compare the performance on two-step and one-step derived forms as well as on novel derivations. One study, for example, showed an auditory lexical decision task, in which legal novel and already existent derivations were displayed in Finnish (Leminen et al., 2010). Elicited responses were measured using electrophysiology. An effect called N400 was reported in both cases, which signifies the subsequent mapping of lexical form onto meaning. This was taken as support for the effective parsing of derivations that are novel.
Another task that is used to study morphological processing is a masked priming lexical decision task, in which the first word of each pair is presented very briefly and preceded by a set of symbols and succeeded by the second word to which a word/nonword lexical decision is made. Such a design allows you to determine whether there are differences between the processing of real morphological pairs (e.g. cleaner-clean), non-morphological pairs (planet-plan) and pseudo-morphological pairs with morpheme-like chunks (proper-prop) (e.g. Rastle et al., 2004). Effects of morphological conditions such as semantically transparent morphological derivations (development-develop) or identical words (table-table) can also be compared to form priming (e.g. scandal-scan) to establish morphological effects.
Morphologically related word pairs elicit either an effect called N250 attenuation or both N250 and N400 attenuation in visual masked priming (Morris et al., 2008). Research that compared between the priming of form-related word pairs, pseudo-derived word pairs and morphologically related word pairs revealed more priming by morphologically related word pairs in the N250 and N400 latency range than by form-related and pseudo-related words (Morris et al., 2007). N400 and N250 effects both reflect the time-course of processing of complex words, specifically, the early stages of lexical processing.
Insights from research into morphological decomposition
What is known from research on morphological processing using such paradigms is that we access morphological units of derived words during the process of word recognition. The question that has been the focus up until recently is when exactly we access and process each unit. It has been considered that there are distinct units that are processed and segmented at different phases of the process of word recognition. Two processes that have been linked to the morphological decomposition of derived words over the course of word recognition are orthographic and semantic processes.
Research in this area has been dealing among others with questions such as whether morphological units such as derivational affixes are either just a by-product of statistically recursive orthographic parts (known as the morpho-orthographic perspective), the outcome of a semantic analysis of information that affects the earliest word recognition stages (known as the morpho-semantic perspective), or if morphological units emerge at a point at which both semantic and orthographic knowledge are made use of at the same time. In the latter case, the processing of morphological units would actively follow both morpho-orthographic and morpho-semantic routes.
The majority of studies on morphological processing support the decompositional dual route perspective in terms of the observed patterns of response and activation (Leminen et al., 2019). The location and latency of the morphological effects, however, vary very much dependent on the linguistic variables that were under investigation, and on the paradigm that was used in each study. Because of the conflicting nature of the results across studies regarding the location of morphological effects, it has been therefore suggested that the processing of derivational complex words entails a network of areas in the brain. This network includes both regions that are particular to the modality in which stimuli are presented, as well as the main language-related fronto-parietal areas of the brain.
It is therefore clear that in order to create a better understanding of derivational processing, future empirical research needs to use paradigms, and stimulus characteristics that are more uniform and that even allows for comparisons to be made across studies that investigate this subject in different languages. It is only then that a completely thorough picture of how at the temporal and spatial level derived words are processed and what precise neural processes underlie morphological decomposition can emerge.
References
Aronoff, M. (1980). Contextuals. Language, 56(4), 744–758.
Gold, B. T., & Rastle, K. (2007). Neural correlates of morphological decomposition during visual word recognition. Journal of Cognitive Neuroscience, 19(12), 1983–1993.
Leminen, A., Leminen, M. M., & Krause, C. M. (2010). Time course of the neural processing of spoken derived words: An event- related potential study. Neuroreport, 21, 948e952.
Leminen, A., Smolka, E., Dunabeitia, J., Pliatsikas, C ,(2010). Morphological processing in the brain: The good (inflection), the bad (derivation) and the ugly (compounding). Cortex, 116, 4-44.
Marslen-Wilson, W. D., Tyler, L. K., Waksler, R., & Older, L. (1994). Morphology and meaning in the English mental lexicon. Psychological Review, 101(1), 3–33.
Morris, J., Frank, T., Grainger, J., & Holcomb, P.,J. (2007). Semantic transparency and masked morphological priming: An ERP investigation. Psychophysiology, 44, 506e521.
Morris, J., Grainger, J., & Holcomb, P.,J. (2008). An electrophysiological investigation of early effects of masked morphological priming. Language and Cognitive Processes, 23, 1021e1056.
Pliatsikas, C., Wheeldon, L., Lahiri, A., & Hansen, P. C. (2014). Processing of zero-derived words in English: An fMRI investigation. Neuropsychologia, 53, 47e53.
Rastle, K., Davis, M. H., & New, B. (2004). The broth in my brother's brothel: Morphoorthographic segmentation in visual word recognition. Psychonomic Bulletin & Review, 11, 1090e1098.
