"When two vowels go walking, the first one does the talking"

[forty-two]

This rule (or "rule") seems to be the go-to example for so-called phonics rules that aren't. "It's only true 43% of the time!" People use it to show why "phonics doesn't work", and phonics advocates are quick to throw it under the bus as an example of crappy phonics teaching.

 

Which is why I was shocked to see that LiPS, an intensive, research-based remedial program, actually teaches it. What in the world?!? But I thought about how LiPS does it - they first teach the primary spelling(s) for the additional vowel sounds beyond the short/long vowels (au/aw, oo, oi/oy, and ou/ow), and only use the "two vowels go walking" rule on other two-vowel phonograms, as a guide to their most *common* sound (not their *only* sound). And when you do that, the accuracy rate goes way up.

 

There are very few big exceptions: namely, eu/ew, which is only used for /OO/ and /U/ (and so is outside the scope of the rule), and "ie", which is the only true exception, because "the second letter does (most of) the talking". There are a few more phonograms where another bit of phonics knowledge is needed to complement the "two vowels" rule (usually wrt foreign spellings). The big one is "ey"/"ei"; the primary sound is the *Latin* sound for "e" (/A/) and it is used most in words of Latin origin; the next most common sound for these phonograms is long-e (and actually, for the suffix -ey, /E/ is the *only* sound). "ui"/"ue" both have second sounds with consistent, easily identified criteria (-ue is silent at the end of words of French origin, and in some words that start with g, the "u" in "ui" is silent, and serves to separate the "g" from the "I", making it clear it's a hard "g").

 

 

So I ran the numbers (using The ABCs and All Their Tricks as the source of all my numbers), and I'm starting to think that this much maligned rule doesn't deserve its bad rap. For the following list of phonograms (basically the Spalding two-vowel phonograms, excluding au/aw, oo, oi/oy, ou/ow, and eu/ew as explained), here's how the "two vowels go walking" rule of thumb works out:

 

When you include the known outlier "ie", it identifies the correct sound 90% of the time.

When you *exclude* the known outlier "ie", it identifies the correct sound 95% of the time.

 

If you don't include second sounds with consistent, easily identified criteria (ue/ui), to look only at true outliers, it identifies the correct sound 98% of the time. I think that's a pretty darn good rule of thumb :). I think a key thing is to see "two vowels go walking" as a *rule of thumb*, a guide to which sound to try first, to the *most common sound* for phonograms that aren't primarily used for non-short/long vowels, instead of as some absolute, "it's *always* this sound" rule. Within those parameters, it's quite accurate and useful :thumbup:.

 

 

Here's my numbers, for the phonics geeks ;):

 

ai: long-a, 98%; other, 2% (long-a: 308; other: 5)

ay: long-a, 99%; other, 1% (long-a: 143; other: 2)

 

ea: long-e, 67%; short-e, 32%; other, 1% (long-e, 325; short-e, 156; other, 5)

ee: long-e, 99%; other, 1% (long-e: 307; other: 2)

ei: Latin long e (long a), 73%; long-e, 15%; other: 12% (Latin long-e: 54; long-e: 11; other: 9)

ey: in base words, Latin long-e, 85%; other, 15% (Latin long-e: 17; other: 3)

as a suffix, long-e, 100% (43, no exceptions)

 

ie (the outlier): within words, long-e, 97%; other, 3% (long-e: 77; other: 2);

at the end of words, long-e, 63%; long-i, 37% (long-e: 17; long-i: 10)

 

oa: long-o, 99%; other, 1% (long-o: 132; other: 1)

oe: long-o, 83%; other, 17% (long-o 15; other: 3)

 

ue: long-u, 51%; silent at the end of words from French, 49% (long-u: 44; silent: 43)

ui: long-u, 62%; silent u, 38% (long-u: 15; silent u: 9)

[Ellie]

Huh.

 

Well, Spalding doesn't teach that as a rule, and I'm glad. :-)

 

Methinks you're more of a math geek than a phonics geek, lol.

[SoCal_Bear]

Hmmm...that's interesting. I am am copying what Wanda Sanseri writes in Spell to Write and Read about this specific rule:

 

"Fickle phonics teachs useless ideas like the cute sounding rule, “When two vowels go walking the first one does the talking.†In other words, if we see two vowels together, the first one will say the name of the letter and the second one will be silent. (OA = /O/). Back in the ‘70s, I had my students mark page after page of words that illustrated this principle. The concept worked on screened worksheets, but in real life it failed repeatedly. I discovered why. The “two vowels going walking†rule is reliable only 27% of the time! It only works consistently with aigh, ee, oa, oe. It commonly works with ay, ai. It possibly works with ea, ae. It usually does not work with ei, ey, ie, oo, ou. It never works with au, augh, ear, eau, eu, oi, oy, ui."

 

 

 

[ElizabethB]

I have run the numbers, I have not only The ABCs but the original Hanna study with all 17,000 words.

 

http://www.thephonicspage.org/Phonics%20Lsns/Resources/letter%20sound%20read%20new%202011.pdf

 

For spelling:

 

http://www.thephonicspage.org/Phonics%20Lsns/Resources/sound%20letter%20spell1.pdf

 

I also have color coded charts.

 

http://www.thephonicspage.org/On%20Reading/Resources/PL26VowelChart.pdf

[fdrinca]

Completely anecdotal, and n=1, but:

 

I started using "two vowels walking" with my daughter, nervously I might add. I knew it didn't hold up over most of the phonograms. Then, I realized that she only struggled with the words that followed this rule. 

 

But I'm comforted by your numbers :)

[ElizabethB]

I also have language of origin charts, within each language of origin, English is actually very phonetic.

 

http://www.thephonicspage.org/On%20Reading/Resources/OriginSlides.pdf

[Piper]

I have run the numbers, I have not only The ABCs but the original Hanna study with all 17,000 words.

 

http://www.thephonicspage.org/Phonics%20Lsns/Resources/letter%20sound%20read%20new%202011.pdf

 

For spelling:

 

http://www.thephonicspage.org/Phonics%20Lsns/Resources/sound%20letter%20spell1.pdf

 

I also have color coded charts.

 

http://www.thephonicspage.org/On%20Reading/Resources/PL26VowelChart.pdf

 

 

I also have language of origin charts, within each language of origin, English is actually very phonetic.

 

http://www.thephonicspage.org/On%20Reading/Resources/OriginSlides.pdf

 

Thanks for posting these links; they are fascinating, especially the last one.  

 

On a personal note, the phonics program I use doesn't teach the "two vowels go walking" rule either.  :-)  I'm very impressed by your analysis,OP!  Thanks!