Cat People (1982)

Readability Formulae: Facts And Limitations

Readability Formulae: Facts and Limitations

Introduction

How easy is it to read a text? How clearly does a text express ideas and emotions? How could the linguistic difficulty of a text be determined and analysed? These questions have been one of the major concerns of text designers, teachers and writers. In actual fact, these questions are inextricably bound up with the concept of readability. In an attempt to provide answers to these questions, so many studies and research have been conducted and as a result, a large number of formulae were proposed to measure the readability of written texts.

In the theoretical part of this paper, I will highlight some of the ways that are used to determine the linguistic difficulties of written texts. Then, I will explain some of the common examples of readability formulae and how they work. Finally, I will shed some light on the limitations and validity of these formulae. In the practical part, I will try to examine the readability of some texts by using the readability formulae mentioned in the first part.

1/ Determining linguistic difficulty of a written text

How difficult a text is to read would appear to be related to whether or not we understand the words in a text, and whether or not these words are put together in an easy‑to‑follow manner. In fact, readability formulae do use these basic criteria and discuss them in terms of vocabulary difficulty and syntactic complexity (Olson, 1984). Vocabulary difficulty refers to the degree to which a text contains words that are unfamiliar and/or difficult to understand. Syntactic complexity refers to the degree to which the sentences in a text have complicated grammatical structures. I shall now discuss these basic criteria.

a) Vocabulary difficulty

Word difficulty is a criterion used by almost all the standard readability formulae. Difficulty is measured either by the presence or the absence of a word on a list of frequently used words, such as Kucera and Francis’s word list (1967), or by word length which is determined by counting the number of syllables in the word. However, both methods appear to have serious limitations, as I show below.

b) Word lists

A number of readability formulae such as Dale‑Chall (Dale and Chall, 1948) use word lists. The assumption underlying the use of word lists to measure the difficulty of vocabulary in a text is that frequently used words will be more familiar and therefore easier to understand. However, in order to have a word list that includes either familiar or difficult words, one must assume that the words in a language remain relatively stable and this is impossible due to the fact that vocabulary tends to change quite rapidly. If we consider the following words from Dale’s 3000 word list (Harrison, 1980, pp. 153‑163): hairpin, maypole, cobbler and washtub, it would appear to us that these words are likely to be unfamiliar to young readers (elementary school children) because they are rarely used (they are out of fashion). On the other hand, some other new words such as download, hyperlink have entered the language and have become extremely familiar to many if not most people. Now, we might say simply that the former words should be taken off the list of familiar words and the latter words should be added instead. Since a number of readability formulae use certain word lists most of which are old, it could be inferred that the results of using these formulae do not really represent or provide the accurate level of readability for the current written texts.

In addition, word lists never take into account that different socio‑cultural groups of the same generation have very different core vocabularies. Nor can they accommodate the fact that different professional groupings can have radically different vocabularies (Kittredge and Lehrberger, 1982). Another problem with word lists is that words frequently have more than one meaning. In such case, what meaning is to be considered a familiar one? It is only through viewing and examining the word in the context of a particular text that one can know about the intended meaning. In conclusion, word list provide neither an indication of which meaning is common (familiar), nor a means to identify the particular meaning that is pertinent in a specific context.

c) Word complexity

In many readability formulae (for example, the Fry graph, the Flesch formula, Gunning Fog formula, see Harrison, 1980), the longer a word is, the more difficult it is assumed it is to comprehend. The idea is applied in different ways. In the case of Flesch and the Fry formula for example, the criterion is the number of syllables per 100 word of text whereas for the Gunning Fog formula, the criterion is the number of polysyllabic words. Regardless of how word complexity is measured by different formulae, however, the assumption is that word length directly relates to the ease or difficulty with which a text can be read.

The problem with this assumption however, is that long words are not always the most difficult to understand. On the contrary, there appear to be a significant number of instances where mono or bisyllabic words are more difficult and unfamiliar than longer polysyllabic words. If we consider the words curr and unemployment, the number of readers who know the latter term is quite likely greater than those who know the former. From the previous example, we could say that word length is often the direct result of affixation (attaching prefixes e.g. un‑ or suffixes ‑ment with shorter words and word stems, as in unemployment). In actual fact, people are likely to be aware of the functions of these affixes and prefixes and therefore, a more complex word ( a word with one or more affixes) may be just as easy to understand as a simple word. However still, reading formulae that count the number of syllables per hundred words would favor a text that has only uninflected forms of verbs.

Furthermore, Randall (1988) performed experiments on the relationship between comprehension and morphological complexity. She tested children between ages of 3 and 7 on their comprehension of morphologically complex terms (nouns derived from verbs by the affixation of ‑er). Her results suggest that the comprehension of complex terms has nothing to do with word length or frequency of occurrence, as would be predicted by readability formulae.

d) Syntactic Complexity

Almost all readability formulae equate sentence length with syntactic complexity. They consider that there is a correlation between the average length of sentences in a text and the difficulty of that text. In fact, closer examination shows that sentence length may actually facilitate comprehension, rather than impede it. For example, the presence of coordinating conjunctions and logical connectives increases the length of a sentence, but these connectives often make explicit its communicative intent. Compare for example, the sentences of (1) with the sentence in (2):

(1).       a. I could not answer your e‑mail.

b. All the computers were occupied.

(2)        a. I could not answer your e‑mail because all the computers were occupied.

In (1), there could be various reasons that the author was unable to respond to an e‑mail aside from the fact that all computers were occupied; he could have not had enough time to wait until one of the computers is free, he could have a lecture or a meeting so he could not wait..etc. The because in (2), on the other hind, explicitly links the author’s inability to answer the e‑mail to the occupying of all computers. In conclusion and based on the above examples, it could be said that equating syntactic complexity with sentence length is neither a useful nor an accurate criterion for measuring readability.

2/ Examples of readability formulae

There is a huge number of readability formulae which have been designed for very different purposes, and for a variety of age groups. These formulae have been designed by psychologists, researchers and classroom teachers and they have been derived using different statistical techniques and measurements. In this part, I will focus only on four of them for two reasons. Firstly, these four are some of the most commonly used readability formulae. Secondly, they are very easy to use and they don’t require complex mathematical or statistical methods. These formulae are as follows:

A) The Flesch Formula

The Flesch formula is one of the very widely used and well‑known readability measures which was designed by Flesch (1948). The formula uses the average number of words per sentences and the average number of syllables per 100 words as variables. Flesch was interested in assessing adult reading material, so he chose a difficulty index that did not relate to grades, but to a notional comprehension score out of 100 (Hlarrison, 1980). When applied to a document, the Flesch formula results in a number ranging from 0 to 100. The lower the score, the more difficult the material is to read and comprehend. Moreover, Flesch provided a transformation table which makes it possible for us to relate the reading ease score to age level. This transformation table is a nomograrn (a figure) which can be easily used by using a ruler between the left‑hand and right‑hand columns of the figure. The Flesch formula is not only used manually, it is also used in most of the word processing programs such as Microsoft word. The only difference is that in the computer adaptation of the Flesch formula, the syllable count is replaced by a vowel count which gives almost the same results as if the count was that of a syllable. Research by Coke & Rothkopf (1970) has showed that counting vowels provided very similar estimates to counting syllables.

B) The Fry Graph

Like most readability formulae, the Fry Graph has a syntactic factor which is the sentence length and a semantic factor which is the number of syllables. The Fry Graph is one of the most straightforward ways of obtaining a readability index. The use of the graph is very useful in this formula for different reasons as stated by Harrison (1980):

1) it saves time on making calculations

2) it offers visual information when numerical results might give a spurious impression of accuracy

3) the user of the graph can tell at a glance if a passage is in comparative term more difficult than average in vocabulary or in sentence length. (Harrison, 1980.p 73)

In the graph, the curve represents normal texts. So clearly, points above the line, or towards the top right quadrant, will represent passages with higher than average vocabulary difficulty. On the other hand, points below the curve, towards the bottom left quadrant, will suggest greater than average sentence length (ibid). To implement the Fry Graph, one should first randomly select three sample passages of exactly 100 words (from the beginning, the middle, and the end of a text). After the total number of sentences and syllables for each of the 100‑words passages has been recorded, the average number of sentences and syllables is computed. The resulting figures are then plotted on the Fry Graph and the resulting coordinate point is associated with an established grade level designation (further explanation of the use will be shown in the practical part). The fry Graph is appropriate for assessing materials from the first grade through the college level (Fry, 1969).

C) The FOG Index Formula

The Fog Index, which uses as few as 100 successive words to determine both sentence length and the number of words with three or more syllables, was developed by Gunning (1968). The counts are then substituted into a formula and the reading difficulty is calculated according to formal grade level in school. For longer written works, the author recommends selecting several 100 word samples from various parts of the material and then counting both the sentence length and the number of words with three or more syllables for all the samples. Finally, the average of the previous counting should be calculated and the Fog Readability Index is found:

Fog Readability Formula: Grade Level = 0.4 * (average sentence length + percentage of words with three or more syllables).

The formula is clearly similar to that of Flesch. The factors (sentence length and number of words) remain the same but the counting of three syllable words will be easier and requires less time than the syllabus count required by the Flesch. A further advantage lies in the simplicity of calculation required to resolve the equation. This formula is appropriate for assessing materials from the fourth grade through the college level.

D) The SMOG Formula

This formula gives the SMOG grade which is the reading grade that a person must have reached if he is to understand fully the text assessed (Gilliland, 1972.p 94). To implement the SMOG formula, one should count 10 consecutive sentences near the beginning, 10 near the middle and 10 near the end of the text. Then, in the 30 selected sentences, the number of words with three or more syllables is counted. Finally, the square root of the number of the polysyllabic words is counted and added to 3. It is clear here that the sentences are used only to prepare the samples; they are not used in the calculation. The SMOG formula gives a readability score of 2 grades higher because McLaughlin used complete success as his criterion for comprehension (ibid). The SMOG is quicker to work out by hand as it does not require counting every word in every sentence but the words with three or more syllables found in 30 sentences.

In the case where a text has less than thirty sentences, McLaughlin (1969) recommended that the number of polysyllabic words and the number of sentences should be counted in the whole text. The process of finding the readability in a text of less than 30 sentences is demonstrated in the practical part (examining the given texts). It is also available in the Appendix.

3/ Factors affecting readability

As we have seen previously, almost all the readability use word complexity, sentence length and syntactic complexity represented by word length or the number of syllables in a word. Many other factors that could affect readability to some great extent have not been considered by these formulae. Here, I will discuss how these factors affect readability and why they should be taken into account when measuring readability. The factors are as follows:

a) Grammar

I have mentioned before that most of the readability formulae tend to equate syntactic complexity with sentence length which is really a very narrow way of looking at complex syntax. By doing this, readability formulae ignore some basic issues such as the conventional use of prescriptive grammar for standard written English. In other words, they ignore the fact that errors or deviations from the standard grammar may make a text more difficult to comprehend and they do not take into account that the presence of a substantial number of grammatical errors can seriously impede the reader’s ability to comprehend the text. If we consider the potential confusion caused by a faulty parallelism such as in:

‑ Not only do we dislike him but also his wife.

It is not really clear whether the intended meaning of this sentence was that we dislike both him and his wife, or that we are not the only ones who dislike him; his wife dislikes him too. There are some other grammatical errors that could be a source of confusion in a written text and some of these are run‑on sentences, sentence fragment, faulty parallelism and pronoun references. In conclusion, we could say that grammatical errors must be considered in the process of measuring the readability of any text in order to make that measurement precise and accurate.

b) Style

Sometimes, a text may be perfectly grammatical, but certain stylistic properties of some text may make them relatively more difficult for people to process than others. One of these stylistic properties is the number of clauses in a sentence and the type of these clauses. In actual fact, there is empirical evidence that certain syntactic properties make some texts easier to process than others. For instance, Van Dijk and Kintsch (1983) note that sentences with relative clauses are easier to parse when a relative pronoun is overtly present than when it is not. Van Dift and Kintsch compare the following sentences and note that people perform better on comprehension tasks in (1) with overt relative pronouns, than in (2), where they are absent:

1) The pen which the author whom the editor liked used was new

2) The pen the author the editor liked used was new (Van Dijk and Kintsch, 1983, p.29)

It is thus clear that the syntactic structure of sentences is not the only stylistic property of a text that can affect readability. The use of figurative language is a non‑stylistic feature that can pose challenges to the reader because understanding figurative language requires more than simply understanding the literal meaning of a text. Readability formulae neither consider the stylistic nor the non‑stylistic features when measuring the readability of a text and this in turn, affects the accuracy of the overall results.

c) Background Knowledge

Readability formulae would judge a text to be easy or difficult by measuring the properties of that text; they never take into account characteristics of readers. These different formulae have been designed for the use with native speaker children and not for second/foreign language learners. No can deny that these groups differ in their background knowledge as well as in their language competence and ability to comprehend written texts. There is evidence suggesting that the characteristics of readers (which have been ignored by readability formulae) have a great deal to do with their ability to comprehend a text. For instance, Barry and Lazarte (1998) report a result of a study on effects of background knowledge and syntactic complexity on recall and inference generation among high school second language learners of Spanish. Their findings suggest a complex interaction between the level of syntactic complexity of the text itself and the prior knowledge of students on the generation of inferences about text. Finally, it could be said that material that is easily read by a group will not be as easy for another and that is partly because learners have different background knowledge about different topics. Therefore, readability formulae should consider this important factor when measuring the readability of any text.

d) Textual coherence

Most of the readability formulae limit the scope of their research to the sentence boundary. Moreover, they evaluate texts in terms of the properties each of their sentences contain. However, the readability of a text is not limited to a function of the average length of its words or the average number or words in its sentences; its readability is rather a function of the connections and interrelationships between the sentences in a text. Bailing and Grafstein (1991), suggest that in order for a text to be­understandable, it must make clear to the reader the basic logical relationships between sentences so that he/she is able to figure out from the text, who is doing what to whom. In other words, the sentences in a text must have logical and sequential relationships in order to make the text easy to comprehend. Finally, Harrison (1986) states that the presence of explicitly‑stated logical connections often makes a contribution to the coherence of a text.

4/ Limitations and validity of readability formulae

Despite the fact that readability formulae are widely used and relied upon in designing textbooks, they have so many limitations which can affect their validity to a certain extent. Readability formulae were originally developed to try to ensure that a school textbook for a particular grade was appropriate for native speaker children at that grade level. However, even within the same grade level, children differ from each other and sometimes these differences are very great to the extent that the same readability formula cannot possibly be adequate for all of them. In other words, readability formulae do not distinguish audiences; they rather assume that all readers are alike. If this is the case with native speaker children, then these formulae are more likely to have more problems when used with second/foreign language learners. Klare (1963,pp. 24‑25) states:

Formulae measure only style. They do not touch on content, organization, word order, format, or imagery. Nor do they take into account the differing purpose, maturity, or intelligence of readers.

Based on this, it becomes clear to us that readability formulae are never meant to say anything about forms or other primarily visual materials although these forms and other visual materials play a big role in enabling the reader to understand the text. In another related study, Klare (1976), reviewed 36 studies that attempted to improve comprehension by improving readability scores. She concluded that improving comprehension does not correlate well with improving readability scores because readability formulae measure only what can be counted in the text. The reason why this experiment didn’t work is that readability formulae only measure the symptoms of readability. Difficult ideas are usually expressed in difficult language (which is appropriate), the ideas won’t become less difficult just because the language is simplified. It is thus clear that judgment about level of text difficulty go beyond the statistical nature of readability formulae that tend to rely on measures of semantic (usually word length or word frequency) and syntactic (usually sentence length) features.

In addition, readability formulae do not consider the fact that there is an interaction between the text and the reader. They do not consider some important factors on the reader’s side such as his knowledge about the text and his motivation to read. All these factors have a great effect on making the text much easier or much difficult to read. The only thing these formulae consider is the text and they ignore all other factors. Thus, they are inadequate because they consider only one source of information which is contained on the printed page.

Since readability formulae are designed to analyse prose text, problems arise when these formulae are applied to technical, scientific, and mathematical materials that contain a great deal of numerical and symbolic language and specialist vocabulary in addition to prose. Of course, being applied to these kinds of texts, readability would continue to use the traditional measures of word length and sentence length and this in turn, would never reflect the actual readability level of these texts. Finally and based on all that mentioned above, we should not rely on readability formulae alone in selecting materials; we should rather seek the opinion of experts or get reliable consensus opinions to examine characteristics that formulae cannot predict. It is only if we incorporate all these factors that readability formulae could be of benefit to us; otherwise we would not really get much out of using them.

5/ Examining the readability of some texts.

In this part, I will apply the previously explained readability formulae in order to measure the readability of two different texts. I will start first by examining Text A (See Appendix 3) using the four formulae and then I will move to examine Text B (See Appendix 4) using the same formulae.

4/  McLaughlin ‘SMOG’ Formula.

Because Text (B) does not have 30 sentences, I have considered the total number of sentences in the text and the total number of words with 3 or more syllables as recommended by McLaughlin (1969), see Appendix (2).

The total number of sentence in Text B = 18

The total number of words with 3 or more syllables = 58

The corresponding conversion number = 1.67

The Grade Level = The total number of words with 3 or more syllables * The conversion number

= 58 * 1.67 = 96.86

The value 96.86 falls between (91 ‑ 110) on The SMOG Conversion Table 1. Therefore, the grade level is 13.

As can be seen in the above table, no two formulae come up with exactly the same grade level or reading age. However, some formulae come up with very close values for both level and age, as in the Gunning FOG and the SMOG formulae. Similarly, Larrick (1954) compared the grade placement indicated by five formulae for two children’s books. He found out that there were considerable variations between the formulae in terms of the grade level. This emphasises the fact that readability can not be accurately measured by merely using readability formulae. The learner’s ability to comprehend, the style, and the rest of the factors that have been mentioned in part one should be taken into account and should never be ignored when measuring readability.

Conclusion

In conclusion, researchers and textbook designers should be aware of the fact that using readability formulae only does not reflect the actual readability level of a text. In addition to these formulae, they should try to incorporate all the other factors that could affect the readability of texts whether these factors are available within the texts (word complexity and sentence length) or outside the texts (factors that are to do with the reader’s knowledge, motivation and the individual differences between readers within the same grade level). They should try to take these factors into account when measuring the readability of written material. Moreover, they should consult reading experts and teachers in an attempt to reach a consensus with regard to the readability of texts and their suitability to the students’ grade level.

Bibliography

Bailing, A and Grafstein, A. (199 1). The assignment of thematic roles in Ojibwa. Linguistics 29, 397‑422.

Barry, S and Lazarte, A.A. (1998). Evidence for mental models: how do prior knowledge, syntactic complexity and reading topic affect inference generation in a recall task for non‑native readers of Spanish? The Modern Language Journal 82 (2), 176‑193.

Coke, E and Rothkopf, E (1970). Note on a simple algorithm for a computer­-produced reading ease score. Journal of Applied Psychology 54, 208‑210.

Fry, E. (1969). The Readability graph validated at primary levels. The Reading Teacher, 22, 534‑538.

Olson, A.V. 1984. Readability Formulae‑ Facts or Fiction. University of Victoria, British Columbia (ERIC Document Reproduction Service No. ED 258 143).

Kuceral, H and Francis, W.N. (1967). Computational Analysis of Present‑Day American English. Province, RI: Brown University Press.

Harrison, C. (1980). Readability in the Classroom. Cambridge: Cambridge University Press.

Harrison, C. (1986). New directions in text research in readability. In Cashdan, Asher (Ed.), Literacy: Teaching and Learning Language Skills. Basil Blackwell, Oxford, pp. 61‑81.

Kitteredge, R and Lehrberger, J . (Eds.), (1982). Sublanguage: Studies of Language in Restricted Domains. Berlin and New York: Walter de Gruyter.

McLaughlin, G. (1969), SMOG grading: A new readability formula. Journal of Reading, 12 (8) 639‑646.

Randall, J.H.(1988). Candlestick‑makers: the problem of morphology in understanding words. In: Davidson, A  and Green, G (Eds.), Linguistic Complexity and Text Comprehension: Readability Issues Reconsidered. L. ErIbaum, Hillsdale, NJ, pp. 223‑245.

Klare, G. R. (1963). The measurement of Readability. Ames, IA: Iowa University Press.

Larrick, N. (1954). Try it on for fit. Library journal, 79, (April), 729‑733.

Van Dijk, T.A and Kintsch, W. (1983). Strategies of Discourse Comprehension. New York: Academic Press.

Gilliland, John. (1972). Readability. London: University of London Press.

Appendix (1)

EDWARD FRY’S READABILITY GRAPH

Appendix (2):

Taken from: McLaughlin, G. (1969), SMOG grading: A new readability formula. Journal of Reading, 12 (8) 639‑646.

Use this formula and SMOG Conversion Table If for material containing less than 30 sentences, but not less than 10 sentences.

1.              Count the total number of sentences in the material.

2.   Find the total number of sentences and the corresponding conversion number in SMOG Conversion Table II.

3.   Multiply the total number of words with 3 or more syllables by the conversion number. Use this number as the word count to find the correct grade level from Table 1.

SMOG Conversion Table 1 SMOG Conversion Table 2

(for longer materials)                                        (use on material with

Appendix (3)

The three sample passages in Text (A)

Sample One:

Pets boost children’s health

Children who have pet animals at home have stronger immune systems and are less likely to take days off school sick, a study suggests.

Researchers at Warwick University in Coventry found that having a cat or dog exposed children to more infections early in life.

However, this exposure boosted their immune systems in the medium term and meant these children attended school more often, on average, than pupils who did not have pets.

The authors said the benefits were most pronounced in children aged between five and eight years.

Dr June McNicholas and colleagues tested the saliva of 138 (eight is excluded).

Sample Two:

High levels of IgA suggest that the immune system is under strain while low levels show that it is vulnerable to infection.

Health benefits

The study showed that antibody levels among pet owning children were significantly more stable, indicating that they had robust immune systems.

Pet owning children were found to have an extra nine days at school over the course of the year compared to those without animals.

According to the researchers, the findings appear to support the so‑called “dirty hypothesis”.

It suggests that too much cleanliness early in life can leave the immune system weakened later on.

Sample Three:

Dr McNicholas, a health psychologist who led the study, said: “Pet ownership was significantly associated with better school attendance rates.

“This was apparent across all classes, but was most pronounced in the lower school (classes one to three, aged groups five to eight)”.

“Here, the pet owners benefited from up to 18 extra half days schooling per annum than their non‑pet owning counterparts.”

However, Dr McNicholas warned that pets can also pose health risks to children.

One of the biggest risks is the roundworm Toxicara canis which infects dogs and can cause anything from stomach ache to eye damage.

Appendix (4):

The Three Sample Passages in Text (B):

Sample One:

Pets And Children: A Lifelong Friendship

Is your voice beginning to sound like a broken record? Are you always nagging your kids to do their homework? Are you having trouble motivating your child to stop watching TV or playing video games and to play outside instead? Then consider making a pet part of your family.

Many people are aware of the health benefits that come from having a pet, includinglowering high blood pressure, preventing heart disease and combating depression. However, what parents may not realize is that adding Spot, Polly or Mr. Whiskers to the family can be advantageous to the other bundles of joy

Sample Two:

While little children are too young to worry about preventing stress or lowering health care costs, there are numerous benefits they can experience from having a family pet.

Pets, whether a dog, cat, bird, hamster, reptile or fish, help children gain a sense of independence that can set them on the path to becoming mature, responsible adults.

Pets Teach Kids To Be Responsible

Children can learn the importance of responsibility at an early age by acting as a caretaker for a pet. Fish are a terrific first pet because children can play a large role in caring for them. However, other pets that require more attention

Sample Three:

Showing children what it means to be responsible for another creature’s survival can result in teaching important life lessons such as discipline, patience, kindness and attentiveness.

Pets Can Help Kids Develop Discipline

Walking the dog, feeding the guinea pig and talking to the parrot can serve as fun study breaks for kids, and a replacement for television programs and video games. These pet­related activities help children remain focused on the task at hand, and are less likely to become distractions that will prevent homework and chores from being completed.

Pets Prepare Kids For

Life Situations

Bringing a pet home and into the family can be an effective way to help prepare

About the Author

Rashid Al Maamari

BA in English for English Specialists from Sultan Qaboos University (2001)

MA in ESP from the University of Warwick (2003)

Teaching English Language in the Language Centre at Sultan Qaboos University since 2001

Office Tel: +968 24142854

Mobile: +968 99378100

E-mail: rashidm@squ.edu.om

Annette O’Toole in Cat People (1982)

Cat People: Original Soundtrack $5.85 No Description Available.Genre: Soundtracks & ScoresMedia Format: Compact DiskRating: Release Date: 17-NOV-1992...

Cat people (1982) / Vinyl single [Vinyl-Single 7''] ...

Cat people (1982) / Vinyl Maxi Single [Vinyl 12''] ...

Cat People [VHS] $7.95 Paul Schrader, the director of American Gigolo, brought a similar kind of sexual chic to this explicit horror movie. A remake of the beautiful, haunting 1942 Cat People, this version takes off from the same idea: that a woman (Nastassja Kinski), a member of a race of feline humans, will revert to her animalistic self when she has sex. Arriving to meet her brother (Malcolm McDowell) in New Orleans,...

Cat People (1982 Film Starring Nastassia Kinski) [VHS Video] $14.95 VHS VIDEO! Cat People! 1982 Film Directed By Paul Schrader! Starring Nastassia Kinski, Malcolm McDowell, John Heard, Annette O'Toole and Ruby Dee. 1989 MCA Home Video Release. Paul Schrader, the director of American Gigolo, brought a similar kind of sexual chic to this explicit horror movie. A remake of the beautiful, haunting 1942 Cat People, this version takes off from the same idea: that a...