Spatial Attention and Negative Priming

Attention is what allows us to focus on the most salient aspects of our environment. Without attention or habituation, our systems would be overloaded from the constant bombardment of stimuli. With attention we are able to concentrate on the things that we want and are able to ignore others. In spatial attention, we are investigating how we are able to attend to or concentrate on relevant stimuli in our environment while ignoring distractors (sic). In an ERP based study of spatial attention and negative priming, subjects were asked to indicate a targets location while ignoring a distractor that was present in unequal proportions. Negative priming occurs when the processing of a stimulus previously ignored is inhindered in terms of error rate or reaction time.

In priming, it is assumed that previous experiences affect future experiences. Whether the stimulus has higher basal activation levels, is better encoded, has raised the overall activation level of a network or has been stored as a memory trace, all are possibilities used to explain the priming phenomenon. It is assumed that these techniques used to employ positive priming techniques are what hinder the system in a negative priming paradigm. Combining the aspect of attention, if a stimulus is to be actively processed and encoded, it takes your attention to complete, and inversely, attention to suppress the environmental noise or distractors. In negative priming it may be this active supression which explains the decline in behavioral responses.

Two leading theories hypothesize different mechanisms to explain the negative priming effect. Inhibition theory uses this supression of the response while retrieval theory supposes that it is old memory traces which interfer with the behavioral response. In order to parse apart these theories, it is necessary to look into the underlying neural behavior during a negative priming experience. Using ERP, one is able to take advantage of well known neural responses (including some that related to decision making) and good temporal resolution.

Blocked Cyclic Naming (Semantic Category Interference)

This effect reflects the finding that for naming, participants are slower when naming pictures out of a sequence of items from the same semantic category than from different categories. The naming of an object or a picture is a multi-step series of mental operations. If naming a dog, you must first recognize it and decide how you want to refer to that object. This part of the process is called conceptual preparation. You must then focus on the particular mental item ("lemma") in you mental lexicon, which is under competition of semantically related items. The operation of lexical selection is next, and you must access the selected item's form information ("phonological code"). You then syllabify the word by composing it syllables from the segments, this is the processs of "phonological encoding." For each of these syllables you access a stored motor instruction ("gestural score"). Finally you execute the gestures with overt speech as output. This is known as the core processing stages of the production of words (Levelt, Roelofs, and Meyer, 1999).

Cumulative interference for word retrieval by prior retrieval of other members of the same semantic category is similar to an inhibition being caused by these related members. An experiement by Howard et al. (2005) indicates that this inhibition effect can only occur if the spoken word production has three properties, competition, priming and semantic activation. Competition can be implemented through eithe lateral inhibition between candidates or a decision criterion. The priming of previously presented items would persist over time, even when other stimuli are inbetween the targets. The stregthening mapping can be from either semantics or lexical units. Shared activation refers to the activation through connected nodes in a semantically related network being activated. In the experiment, the investigators were able to examine the extent to which there are cumulative effects of semantic competitor priming.. The picture naming latency is slowed by an additional 30 ms for each preceding semantial

Maess, et al. (2002) attempted to distinguish the neuronal correlates of lexical access by using the phenomenon of blocked cyclic naming and MEG recording to note where in the brain, this effect took place. Results demonstrated that the left temporal region supports the processes underlying the semantic interference effect. A previous investigation had demonstrated that the time window of between 150-275 msec post picture onset was the core process of lemma selection. The time window of the current study for the activation of the left temporal region fits this same time frame.

So what does this all mean? In the present study that I am investigating, subjects were asked to name 400 pictures. They were also participants in an auditory lexical decision task in which they were to decide whether a word they heard was a real word or not. I coded these words into semantic categories. In the 400 word naming project there were 5 semantic categories, 3 large ones (animals, tools, and food)and 2 well defined but smaller ones (body parts and instruments). The reaction times of the participants were than regressed against the order of the items in the specific semantic categories. In the presence of other word variables the order of the categories only had an effect in the one group of participants. When the order of the semantic categories is looked at by itself, there is a huge effect of order, demonstrated by a 1.5 ms slowing for successive trials.

This is an interesting effect because you would think that the higher activation of the related targets would lead to a faster reaction time. However, this activation appears to be competitive and inhibitory. Perhaps it is the fact that when a word is being searched out, it has to enter the "search mechanism" with no previous activation. So if I name tiger which activates lion and then am asked to name lion, my mental lexicon has to actually go through a process of finding lion, seeing that it is activated, deactivating it and then go through the normal process.