he may chance to cut a poor figure in the eyes of posterity , for a work which was mere commercial trash to the conoscenti of one generation might possibly become a classic to those of another . if , on the other hand , he is guided by a contempt for the readers of such books , then he is making a crude and unacknowledged use of my system . it would be safer to admit what he was doing and do it better ; make sure that his contempt had in it no admixture of merely social snobbery or intellectual priggery . my proposed system works in the open . if we can not observe the reading habits of those who buy the westerns , or do n&apos;t think it worth while to try , we say nothing about the books . if we can , there is usually not much difficulty in assigning those habits either to the unliterary or the literary class . if we find that a book is usually read in one way , still more if we never find that it is read in the other , we have a prima facie case for thinking it bad . if on the other hand we found even one reader to whom the cheap little book with its double columns and the lurid daub on its cover had been a lifelong delight , who had read and reread it , who would notice , and object , if a single word were changed , then , however little we could see in it ourselves and however it was despised by our friends and colleagues , we should not dare to put it beyond the pale . how risky the current method can be , I have some reason to know . science-fiction is a literary province I used to visit fairly often ; if I now visit it seldom , that is not because my taste has improved but because the province has changed , being now covered with new building estates , in a style I do n&apos;t care for . but in the good old days I noticed that whenever critics said anything about it , they betrayed great ignorance . they talked as if it were a homogeneous genre . but it is not , in the literary sense , a genre at all . there is nothing common to all who write it except the use of a particular machine . some of the writers are of the family of Jules Verne and are primarily interested in technology . some use the machine simply for literary fantasy and produce what is essentially Ma&quot;rchen or myth . a great many use it for satire ; nearly all the most pungent American criticism of the American way of life takes this form , and would at once be denounced as un-American if it ventured into any other . and finally , there is the great mass of hacks who merely cashed in on the boom in science-fiction and used remote planets or even galaxies as the backcloth for spy-stories or love-stories which might as well or better have been located in Whitechapel or the Bronx . and as the stories differ in kind , so of course do their readers . you can , if you wish , class all science-fiction together ; but it is about as perceptive as classing the works of Ballantyne , Conrad and W W Jacobs together as the sea-story and then criticising that . but it is when we come to the second distinction , that made among the sheep or within the pale , that my system would differ most sharply from the established one . for the established system , the difference between distinctions within the pale and that primary distinction which draws the pale itself , can only be one of degree . Milton is bad and Patience Strong is worse ; Dickens ( most of him ) is bad and Edgar Wallace is worse . my taste is bad because I like Scott and Stevenson ; the taste of those who like E R Burroughs is worse . but the system I propose would draw a distinction not of degree but of kind between readings . all the words - taste , liking , enjoyment - bear different meanings as applied to the unliterary and to me . there is no evidence that anyone has ever reacted to Edgar Wallace as I react to Stevenson . in that way , the judgement that someone is unliterary is like the judgement this man is not in love , whereas the judgement that my taste is bad is more like this man is in love , but with a frightful woman . and just as the mere fact that a man of sense and breeding loves a woman we dislike properly and inevitably makes us consider her again and look for , and sometimes find , something in her we had not noticed before , so , in my system , the very fact that people , or even any one person , can well and truly read , and love for a lifetime , a book we had thought bad , will raise the suspicion that it can not really be as bad as we thought . sometimes , to be sure , our friend &apos;s mistress remains in our eyes so plain , stupid and disagreeable that we can attribute his love only to the irrational and mysterious behaviour of hormones ; similarly , the book he likes may continue to seem so bad that we have to attribute his liking to some early association or other psychological accident . but we must , and should , remain uncertain . always , there may be something in it that we can n&apos;t see . the prima facie probability that anything which has ever been truly read and obstinately loved by any reader has some virtue in it is overwhelming . to condemn such a book is therefore , on my system , a very serious matter . our condemnation is never quite final . the question could always without absurdity be re-opened . and here , I suggest , the proposed system is the more realistic . for , whatever we say , we are all aware in a cool hour that the distinctions within the pale are far more precarious than the location of the pale itself , and that nothing whatever is gained by disguising the fact . when whistling to keep our spirits up , we may say that we are as certain of Tennyson &apos;s inferiority to Wordsworth as of Edgar Wallace &apos;s to Balzac . when heated with controversy you may say that my taste in liking Milton is merely a milder instance of the same sort of badness we attribute to the taste that likes the comics . we can say these things but no sane man quite fully believes them . the distinctions we draw between better and worse within the pale are not at all like that between trash and real literature . they all depend on precarious and reversible judgements . the proposed system frankly acknowledges this . it admits from the outset that there can be no question of totally and finally debunking or exposing any author who has for some time been well inside the pale . we start from the assumption that whatever has been found good by those who really and truly read probably is good . all probability is against those who attack . and all they can hope to do is to persuade people that it is less good than they think ; freely confessing that even this assessment may presently be set aside . thus one result of my system would be to silence the type of critic for whom all the great names in English literature - except for the half dozen protected by the momentary critical establishment - are as so many lamp-posts for a dog . and this I consider a good thing . these dethronements are a great waste of energy . their acrimony produces heat at the expense of light . they do not improve anyone &apos;s capacity for good reading . the real way of mending a man &apos;s taste is not to denigrate his present favourites but to teach him how to enjoy something better . such are the advantages I think we might hope from basing our criticism of books on our criticism of reading . but we have so far pictured the system working ideally and ignored the snags . in practice we shall have to be content with something less . the most obvious objection to judging books by the way they are read is the fact that the same book may be read in different ways . we all know that certain passages in good fiction and good poetry are used by some readers , chiefly schoolboys , as pornography ; and now that Lawrence is coming out in paperbacks , the pictures on their covers and the company they keep on the station bookstalls show very clearly what sort of sales , and therefore what sort of reading , the booksellers anticipate . we must , therefore , say that what damns a book is not the existence of bad readings but the absence of good ones . ideally , we should like to define a good book as one which permits , invites , or compels good reading . but we shall have to make do with permits and invites . there may indeed be books which compel a good reading in the sense that no one who reads in the wrong way would be likely to get through more than a few of their pages . if you took up Samson Agonistes , Rasselas , or urn burial to pass the time , or for excitement , or as an aid to egoistic castle-building you would soon put it down . but books which thus resist bad reading are not necessarily better than books which do not . it is , logically , an accident that some beauties can , and others can not , be abused . as for invites , invitation admits of degrees . permits is therefore our sheet-anchor . the ideally bad book is the one of which a good reading is impossible . the words in which it exists will not bear close attention , and what they communicate offers you nothing unless you are prepared either for mere thrills or for flattering daydreams . but invitation comes into our conception of a good book . it is not enough that attentive and obedient reading should be barely possible if we try hard enough . the author must not leave us to do all the work . he must show , and pretty quickly , that his writing deserves , because it rewards , alert and disciplined reading . it will also be objected that to take our stand upon readings rather than books is to turn from the known to the unknowable . the books , after all , are obtainable and we can inspect them for ourselves ; what can we really know about other people &apos;s ways of reading ? but this objection is not so formidable as it sounds . the judgement of readings , as I have already said , is twofold . first , we put some readers outside the pale as unliterary ; then we distinguish better and worse tastes within the pale . when we are doing the first , the readers themselves will give us no conscious assistance . they do not talk about reading and would be inarticulate if they tried to . but in their case external observation is perfectly easy . where reading plays a very small part in the total life and every book is tossed aside like an old newspaper the moment it has been used , unliterary reading can be diagnosed with certainty . where there is passionate and constant love of a book and rereading , then , however bad we think the book and however immature or uneducated we think the reader , it can not . ( by rereading I mean , of course , rereading for choice . a lonely child in a house where there are few books or a ship &apos;s officer on a long voyage may be driven to reread anything faute de mieux . ) when we are making the second distinction - approving or censuring the tastes of those who are obviously literary - the test by external observation fails us . but to compensate for that , we are now dealing with articulate people . they will talk , and even write , about their favourite books . they will sometimes explicitly tell us , and more often unintentionally reveal , the sort of pleasure they take in them and the sort of reading it implies . we can thus often judge , not with certainty but with great probability , who has received Lawrence on his literary merits and who is primarily attracted by the imago of rebel or poor boy makes good ; who loves Dante as a poet and who loves him as a Thomist ; who seeks in an author the enlargement of his mental being and who seeks only the enlargement of his self-esteem . 