Do humans recognize word shapes or letter shapes in order to read?

[MadderMitch]

According to this guy, it appears that we read using a type of parallel letter recognition technique.

As far as I can make out, we basically process the shape of each individual letter in a word at the same time, and therefore determine the word itself. We then derive the semantics of the word using “back up” files in the brain.

This is different from the word shape recognition technique, which was espoused some time earlier. According to this theory, we determine a word by recognizing the shape of the whole word first. Then we go on to process its meaning.

The former theory sounds a little odd to me. I have been speed-reading for some years now, and although most of it is really just a ruse for poor skimming, one technique that I have been taught that has worked to improve both speed and comprehension has been to include whole words or sequences of words in one big chunk.

This means that when I read, I take in the shape of the whole word first (at least, according to me), and then I go about allocating it’s meanings.
So for example, the sentence:

It was a hot day, and Alice went for a walk in the woods
in my interpretation would be-
It was a hot day {Chunk1}

and Alice went for {Chunk2}

a walk in the woods {Chunk3}
Now I’m really confused as to whether I do this or not at all.

My objective is to read as fast as possible with no sacrifice to comprehension, however according to this guy it seems that’s not how humans understand at all.

Even intuitively to me it seems like we recognize words as opposed to letters.

How much evidence is there for either case?

[daffyduck]

You may have seen this:

I cdnuolt blveiee taht I cluod aulaclty uesdnatnrd waht I was rdgnieg The phaonmneal pweor of the hmuan mnid Aoccdrnig to a rscheearch at Cmabrigde Uinervtisy, it deosn’t mttaer inwaht oredr the ltteers in a wrod are, the olny iprmoatnt tihng is taht the frist and lsat ltteer be in the rghit pclae. The rset can be a taotl mses and you can sitll raed it wouthit a porbelm. Tihs is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the wrod as a wlohe. Amzanig huh? yaeh and I awlyas thought slpeling was ipmorantt! so mcuh for sleplscchek!

Here’s what someone working in the Cognition and Brain Sciences Unit at Cambridge has to say about it:

Psycholingusitics

[Cat_Jones]

We touched on this when I was studying for my Diploma in EFL, at that time - a good 7 or 8 years ago - there were two parallel theories, they also applied to how we listen.

Bottom up - where by you look at the letters / hear individual sounds and build up the word, then sentence and work out the meaning.

Top down - which is essentially what you do with texts like the one daffyduck had given us - you recognise essentiel elements and ‘guess’ or ‘fill in’ the meaning.

IIRC mother tongue speakers and those with a good working knowledge of a language tend to use Top Down whereas people with a lower level of language competence (or reading skills) will tend to use Bottom Up - there is a point at which most people we change over. Top down is more efficient ( especially where listening is concerned humans being able to produce more sounds per minute than we can process).

If you have ever tried to learn a new language, especially one with a different script you will know that your reading style is different, initially at least, from the way you read English. I guess it’s a question of training your eye and familiarity; for example with Polish I sometimes have to stop and remind myself how a particular ‘accented’ letter sounded before I could say the word in my head and ‘know’ it. At this stage I approach a text in French as I would one in English. As part of teaching we try to train students to transfer their reading skills - skimming texts, scanning, reading for gist etc. and push them towards a Top down approach.

[smiling_bandit]

To expand on Cat Jones’ comments, the problem with the so-called 'Whole Language" approach is that is tries to skip a step in the learning process. A small number of people may do better with it, but it is not a good idea to try and teach it like that. The phonetic approach recognizes how young people learna nd are likely to work through things. The letters are there in order for a reason.

[Cat_Jones]

[wee clarification] The ‘students’ I deal with are all literate adults learning English for their jobs hence the comment about ‘pushing them’ towards a certain reading style, it’s really a question of getting them to see a text in English as no more problematic than one in their mother tongue and adapting similar strategies - I have no experience teaching youngsters to read for the first time & wasn’t commenting on possible methodology for doing so. Apologies if my post was misleading in that respect. [/wee clarification]

[groman]

Who says that there is one, or even a finite way of reading? Yes, the ways everybody reads might share some similarities, but I don’t think you can just assume that there is one way people read. That would, maybe not imply, but at least hint on the fact that our brains are designed for reading, which is rather a silly concept.

Somebody might recognize letters, others words, others a mix, others might only look at certain parts of the letter before analyzing the whole, etc.

[Hombre]

Deos gliloza isnejt ledias wlohe?

I’m skepical about those mixed up word things.

[Reeder]

Nah…

We learn by rote.

[GuanoLad]

The mixed up word things only work if we already are familiar with the correctly spelled word in the first place.

I’m not sure what relevance that has with the OP necessarily, but it’s something that’s bothrered me about it since I first saw it.

[ParentalAdvisory]

@GuanoLad

The mixed up word things only work if we already are familiar with the correctly spelled word in the first place.

I’m not sure what relevance that has with the OP necessarily, but it’s something that’s bothrered me about it since I first saw it.

Not that it’s correct or not, but it probably has something to do with this bit of the OP:

@MadderMitch

According to this theory, we determine a word by recognizing the shape of the whole word first. Then we go on to process its meaning.

[VernWinterbottom]

Part of our ability to comprehend text quickly is because of familiarity with how the words look. This is why, according to the typography classes I took 25 years ago, it is easier for us to read words in lowercase than it is to read words typed in all caps. The distinctive look of the ascending and descending letters scans more quickly. When all the tops and bottoms of the letters are the same height, we have to work harder to recognize the words.

[MadderMitch]

@VernWinterbottom

Part of our ability to comprehend text quickly is because of familiarity with how the words look. This is why, according to the typography classes I took 25 years ago, it is easier for us to read words in lowercase than it is to read words typed in all caps. The distinctive look of the ascending and descending letters scans more quickly. When all the tops and bottoms of the letters are the same height, we have to work harder to recognize the words.

Actually the uppercase/lowercase argument is mentioned in the linked article in the OP.

Apparently it supports the word-shape recognition model, however:

The weakest evidence in support of word shape is that lowercase text is read faster than uppercase text. This is entirely a practice effect. Most readers spend the bulk of their time reading lowercase text and are therefore more proficient at it. When readers are forced to read large quantities of uppercase text, their reading speed will eventually increase to the rate of lowercase text. Even text oriented as if you were seeing it in a mirror will quickly increase in reading speed with practice (Kolers & Perkins, 1975).

[VernWinterbottom]

Originally Posted by VernWinterbottom
Part of our ability to comprehend text quickly is because of familiarity with how the words look.

This is entirely a practice effect. Most readers spend the bulk of their time reading lowercase text

Right. That’s what I was taught. This is why so many corporate logos and advertising headlines (created by professionals) are in upper and lower case. Amateurs often think printing all their copy in caps will make it stand out, but it’s really harder to read.

In any case, I think most people quickly learn to recognize the whole words. Observation of my five year old daughter also supports this, too. Words that she’s never seen before must be read letter by letter and sounded out. But once she’s seen a word a couple of times, she’ll read it aloud instantly in a book, on a sign, anywhere she sees it.

[Ainak]

1. Notice that when regulations force manufacturers to put ‘warning labels’ or unpleasant notices on their products, the manufactures ALWAYS put them in capital letters, ostensibly to show that they’re making the information prominent. But in reality, they are making it damn near impossible to read. Bold condensed caps is illegible.

2. I’m looking for a cite, but I have seen a rather dramatic demonstration showing that you can read a page perfectly well simply if you only see the tops of the words. Slide off the bottom half, and your eyes see the words just fine.

[Biffy_the_Elephant_Shrew]

@Hombre

Deos gliloza isnejt ledias wlohe?

Correction:

Deos Glilozda isnegt ledias wlohe?

[TheSquirrelfish]

I tend to agree with the whole word recognition theory, and especially the “chunking” part. It is a matter of practice, but eventually you start to recognize what you will read. You see the beginning of a phrase and then only have to check that it continues to follow the pattern you know rather than process each part individually. For example you see “It was a dark” and you might just check to see if something like stormy or night are somewhere in the follow up. This can trick you if something like right or shorey(yeah it’s not a word, but that’s tricky isn’t it) are there instead.

No cites or anything, but my experience as a speed reader. I also know, and recognized this before I’d read any of the theories, that the shapes of phrases and a general recall to the phyisical form of a word are a basic part of reading to me.

[Conan_the_Librarian]

@Biffy the Elephant Shrew

Correction:

Deos Glilozda isnegt ledias wlohe?

I have to agree with Biffy here. The first text given didn’t give me any trouble, but this line did. I couldn’t figure out what gliloza was. Gorrilla? Gonzola? Gill-ooze? When I figured out you left out a word, and it was really Godzilla, then the rest of the line went fairly quickly, even if ingest was incorrect too.
I have found that if you learn a second language similiar to your mother toung, the Top-down method can mess you up sometimes. For example comprar, spanish for “to buy”, had me thinking it meant “to compare”. Luckily these problems aren’t signifigant.

[awldune]

ledias = ideals?

Wouldn’t ‘ideas’ make more sense?

[Cat_Jones]

wrong anagram - ladies

[Excalibre]

@groman

Who says that there is one, or even a finite way of reading? Yes, the ways everybody reads might share some similarities, but I don’t think you can just assume that there is one way people read. That would, maybe not imply, but at least hint on the fact that our brains are designed for reading, which is rather a silly concept.

Somebody might recognize letters, others words, others a mix, others might only look at certain parts of the letter before analyzing the whole, etc.

Why would you think multiple ways exist to perform the same task? That strikes me as considerably less likely than everyone performing the same automatic task in essentially the same way.