Iser’s phenomenological theory of reader response is, to my mind, a little rickety. His main approach centres on what he calls a ‘virtual’ work - an almost living phenomenon - created in unison between an actual text and a reader’s imagination. The virtual creation varies in quality dependant upon the amount of ‘play’ the actual text allows any given imagination. Iser describes the process as being similar to giving names to stellar constellations - the enjoyment derives from using imagination to link up the points into a coherent pattern - too many points and the exercise loses its interest as the pattern becomes too obvious. Iser’s ‘points’ are the expectations of readers to find closure, or at least a semantic flowing of the text, as each sentence follows the next. The failure to find either as the text semantically twists and turns creates ‘gaps’, for which the reader must use their imagination to fulfil the gestalten impulse. It is this process of gestalt that creates the virtual work out of the actual text, and the amount of imagination required by the reader to achieve gestalt which measures the quality of the total experience of the virtual creation. An example might be:

This is an Iserian bad text, as expectation is satisfied and gestalt is totally fulfilled - little imagination is required. An Iserian good text therefore might be:

Here, there is no description of who she marries - expectation of sentence closure is denied (there is no ‘married to’). In the second sentence the pronoun ‘He’ might be anybody - the princess’ father died of a broken heart because of her marriage choice? Expectation of a happy ending is possibly denied by death. Closure is denied as the children must now learn to cope with death in the family, and that is a possible story in itself. Other areas of question exist in this example, but the point is made: expectation is denied, details are missing, a lot of imagination is required to achieve gestalt - resulting in a more vivid virtual work.

As he progresses through the text the reader develops the virtual creation. Failure to realise his expectations, and gaps, force him to revise this virtual creation and this leads to more complex expectations. At each point of call on the imagination, the reader must make choices to fill in the gaps. Each choice being unique to the one particular reader and none other.

This is where Iser’s arguments become weak and woolly. He implies the choice of connections are completely at the reader’s discretion, but is this really the case? For an alien or a computer it might be almost true, but for a reader programmed by his culture text becomes more than just raw material, it becomes stilted. Language is not as dead and dry as Iser seems to claim. Though selection of various connections is wide (dependant on mood, experience, etc), it is not as unlimited as he suggests. Loaded sentences prescribe choice. If Iser denies this effect, he does admit that existence of, or the authorial utilisation of, illusion may prescribe choice in the filling of gaps. In my example above, the term ‘Once upon a time’ is that normally used for fairy stories. This conjures illusions of perfection. Following on from this is ‘a young princess married’, enforcing the illusion and promising a happy marriage to a young prince. The rest of the story disappoints (but not negates) the expectation of her partner’s status, and the deaths at the end shatters the fairy-tale illusion. The reader’s choice of who the princess married is prescribed at the beginning by the illusion to either (for the sake of simplicity) a young prince or evil baron. At the end of the story, at the shattering of the illusion, the choices of partner become much wider (a carpenter would not be impossible). Iser finally makes the point that as text is at the mercy of the reader’s imagination, good text can be identified by what choices it allows the reader to make, and how such choices act to make the reader aware of both his own unconscious impulses, and the life-like experiences he has gained from the text.

To Iser then, the meaning of a text (if any) is unimportant. It is the process of looking for meaning, ‘gestalt’, which is primary, for in doing this a reader gives himself away by how he makes sense of the unwritten. By trying to second-guess alien characters in alien situations he (reputedly) gains insight into both the real-world, and his real-self. But if each experience is different from reader to reader, how might one individual reader say how one text will effect another reader? How can a critic evaluate a text except in an impressionistic and relativistic way? Fish on the other hand approaches text in a more objective way - maybe too objective.

"When analysing a text," Fish claims, "the question is not what it means, but what it does". What a text does provides the answer to what it means. By examining a text word by word, continually scrutinizing how the sentence structure develops over time, linguistic analysis provides the stylistic effect - and this in turn reveals the authorial intent, and thus its meaning. Unlike Iser, whos dead text simply feeds the uniquely individual imagination, Fish suggests that lexical items work to recreate the authorial mind, so that the reader’s experience equates with that of the author’s at time of writing. The difference between Iser and Fish, then, is that one concentrates on deep structure, and the latter on surface structure. Both acknowledge the possibility of multiple interpretations, but a Fishian reader would only interpret differently from another if he was less or more well-informed. Also, differences of opinion arise (between equally well-informed readers) only in how they react to the same interpretation. One may feel pleasure from the stylistic experience, while another may feel impatience. The problem with Fish’s theory is almost diametrically opposed to Iser’s: it is not inconceivable that a computer could be programmed to churn out the meanings of any texts, if properly programmed (with a ‘well-informed’ database). However, considering their differences in the structure depth of their analysis theories, such oppositions may not be mutually incompatible. The gaps of which Iser bases his theory are gaps in the deep structure. Stylistic devices, and deviant lexical positioning rely less on imagination to solve, and more on linguistic competence. Another problem with Fish, is that - though not as extreme as Iser - he still does not do proper justice to the culturally and psychologically ‘loaded’ potentiality of language.

Holland’s psychological theory is by far the most convincing, and most appealing to common sense. The thrust of his argument is that a text is percieved by a reader as a living entity, and responded to as such. Where ‘self’ is created by identity (past experiences, psychological profile, interests, attitudes, etc), a text is created by its ‘unity’ (topic, plots, approach, setting etc). "unity is to text as identity is to self. Or unity relates to identity as text relates to self" he claims. Readers relate to texts in a similar way people relate to people. A person’s ‘perceived’ identity theme is the primary object for immediate character judgements. This judgement affects our response to them - how we initially relate to them - from the first encounter. Further exposure reveals deeper understanding of their identity, resulting in a re-appraisal of their identity theme, resulting in revised judgements. the initial perception of identity - coupled with perception of our own identity - will effect what elements of ourselves we draw upon in order to create a communicative process with them. Where text is concerned, Holland states the same actions apply, with unity theme substituted for identity theme. The process of interacting with the text is entirely subjective, and just as prone to prejudice as is human-human interaction.

The most appealing aspect of this theory is that prejudice plays such a major factor. He claims that the reader - having found some apparent aspect of appeal in his initial judgement - then goes about to map onto the text certain elements of his own identity. He looks for reinforcement of his own identity (or at least aspects it) in the identity (unity) of that read, locking onto some aspects and overlooking others, as judgement permits. He gives an example of one reader’s judgement of Hamlet being centred around the princes struggle against authority, as it matched that reader’s identity theme. Another example is that of three people giving different interpretations of the word ‘fathered’ in a text, based on what that particular word conjured to their personal psyches (heroic, abstract and the third: sexual). Neither Iser or Fish take account of this psychological phenomenon.

Another aspect of appeal in this theory is its ability to take-in all aspects of literature: deep structure, surface structure, form, stylistic, linguistics. Everything which makes up a text can be included. Unity does not gave to mean all aspects are oriented in the same direction, as not all personal experiences and attitudes which make up personal identity are oriented similarly. conflict, and paradox is a natural state, but they are resolved in the psyche - though resolution may shift over time. Similarly, a text may contain conflict and apparent paradox, and these are resolved too by the reader, as, he claims, only those aspects which satisfy and re-enforce the reader’s (current) perception of his own identity. Holland expands his description of the way a reader responds to a text by suggesting that fantasy plays a large role in the acceptance and interpretation of it. Like a Freudian dream, the text allows the reader to create a fantasy from itself, in which submerged desires may be acted out. With this, it may be described that Hamlet acts out the human Oedipus complex for a reader. I personally do not altogether agree with this idea.

Of the three theories examined, Holland’s is the most interesting, as he seems to describe his reader as a human, and not an alien or a computer as seems to be the case for Iser and Fish. Holland also gives full credit to the psychological and cultural power of language. And if any critical theory of literature is to have any credence, this aspect cannot be overlooked, denied, or swept under some linguistic carpet. Humans feel, they associate, they are influenced, they are self-obsessed and they protect their egos and prejudices. All of these aspects of humanity are open to manipulation by an author. To claim the author or the text have little effect on a reader response is untenable. Iser claims a reader is free to choose from a virtually unlimited combinations of links when making sense of textual or semantic gaps. Holland reduces this range of choice by suggesting people are influence by weighted words. Fish suggest words are simply lexical items to be examined - Holland suggests readers accept some words are given more emphasis than others. This is not to reject Fish’s or Iser’s theories out of hand. Both contain important and relevant aspects to understanding the reading process. But only Holland, I believe, gives a more universal and generally acceptable explanation.

Tompkins JP, Reader Response Criticism, Johns Hopkins University Press, London, 1980

Barthes R. Image Music, Text, Fontana Press, London, 1977

Iser W. The Act of Reading, Routlege & Kegan Paul, London, 1978