My Fair Founder
Can master symbolicist Henry Higgins (played by Paul Graham in his first musical-theater role) win a bet by coaching ambitious but crude-speaking Eliza Doolittle to be the toast of Sand Hill Road in three short months?
You'll laugh, you'll cry, you'll sing along to tunes such as:
The Gain in this Domain stays Mainly with the Brain
I've Grown Accustomed to Her Viral Growth Pace
"Why Can't The English Learn To Speak"
As I've posted before, there is this sort of inherent imperialist bias people have with someone who does not have a good command or English. A hundred years ago, it may have been an indication of class (and to some degree still is today). But to look down and be utterly dismissive of someone because they have an accent is just plain dumb.
In PGs defense, just because someone is dumb doesn't mean people will cease to engage in such behavior. And given that the entrepreneur is the one on the selling side, it may behoove him/her to brush up on whichever accent is appealing to those to be sold.
I personally have several accents in English and I subconsciously switch from one to another depending on whom I'm speaking with. If necessary I change languages all together. Reality is, I am better able to connect with the other person this way.
I think that's true in any language. I see this in myself sometimes, and I have to consciously correct myself that one's accent or grasp of a foreign language has nothing to do with intelligence or any other quality other than time spent practicing.
> I personally have several accents in English and I subconsciously switch from one to another depending on whom I'm speaking with.
Not being a native English speaker, I do this too. My American accent is just as fake as my English one, but I think people are used to you speaking a certain accent, and switching seems fake to them. I'll have to start talking to the next Englishman I meet with an English accent, and see how they react when I switch to an American one, which is what I usually use.
* "Can I have a spaghetti" isn't the right way to say it. Drop the "a"
* Have a native english speaker read it out, not a computer generated voice (if that isn't already the plan)
Actual human speech is on the roadmap but not until I fix the main issue, which is finding a repeatable marketing/sales cycle. Open to suggestions though!
This app is mainly aimed at fixing the incorrect pronunciation of words, rather than the accent. If I ever get to add real human voices, then it will be even better.
I will change it to be available for anonymous visitors. Thanks!
Is withholding the price that important that you're willing to risk a reputation black mark on your product?
I just a deployed a fix for this issue, you can now see the prices at http://www.sayafter.me/subscribe
You may find things are more successful with something more like, for example:
Pay Monthly | Pay Quarterly | Pay Yearly
$30/mo | $75 ($25/mo) | $250 ($20.83/mo)
All plans come with a free 30-day trial (or money-back guarantee); cancel or upgrade at any time.
$70 for 3 months
$250 for 12 months.
Since we're on the subject of what your software does, and since as a native English speaker, I don't want to [ab]use your free trial just to study its workings, could you please provide that explanation here? :) Do you teach the basics of English phonology or do you just provide [auto-generated] text for users to mimic. Do you provide Rosetta Stone-like pronunciation feedback? If you don't provide basic instruction in English phonology, I highly recommend you consider adding it. Being well versed in articulatory phonetics, I taught myself Chinese phonology (Standard Mandarin) from Wikipedia. Despite not having very good Chinese, I've passed for a native speaker in short telephone conversations on multiple occasions. Even if it takes weeks to give the student the phonetic background that I had (I don't know if it would), you'll then be able to give them near-native pronunciation almost instantly, and then they'll only need to iron out the kinks with your mimic drills before sounding like a native. I don't know if this is the approach used by most accent coaches, but it's highly effective.
The only feedback the user gets is a score for the current phrase, and highlighting of the words that he has mispronounced.
For example: If the phrase was "Where is the closest hospital?" but you said "Where is the closest clinic?", the application would highlight correct words in green and incorrect ones (hospital in this case) in orange/red depending on the word distance of "clinic" to "hospital")
When I say "word distance" I mean Levenshtein Distance: http://en.wikipedia.org/wiki/Levenshtein_distance
I never thought about giving basic information about English phonology, but you make a very valid point. I will look into this further, and please let me know about any detailed resources about this subject.
Or there's always the simple english version of Wikipedia:
The original "Up Goer Five":
Speaking of which... you could gamify this in like 200 different ways.
Unfortunately currently neither webkit-speech-input nor Web Speech API supports recording of the user voice input.
I will look for alternative solutions though.
But $30/month is more that what I would give. For that price, I would expect to have a human teacher to review my progression from time to time and provide advices.
At $20/month, I would use the service for something like three or four month, until I decide I'm ready enough.
At $10/month, I would probably consider a permanent subscription.
That's less than half what it'd cost to hire a language tutor for a single hour.
I will very likely pull prices down with the next release.
No such thing exists really, at least not in the spoken word form.
You'll also find that regional accents are considered more favourable/likeable than 'the Queens English' so the premise of this may not be 100% correct...
(Brummie is not included in the favourable dialects because it's not considered nice by anyone in the UK except those in that area.)
I know they don't sound ideal, but this is the only distinction I can make via the resources I have.
You made my day, Sr.
It would fall under the 'Queens English' or 'received pronunciation', which, in some studies, is considered less intelligent than Yorkshire dialect - http://www.theguardian.com/uk/2008/apr/04/6
I'm from Yorkshire originally, my dialect is totally different from those in Merseyside or the North East (for example.) How we speak and how we write is different. 'British English' only exists in the written form and I don't believe that is what this service is aiming to 'improve'.
UK regional accents/dialects are something which are favourable. The North East dialect, for example, is seen as favourable in call centres because the majority find it a 'nice' or 'friendly' dialect.
I don't understand the idea of 'if you're a start-up don't have an accent/dialect'.
(Noting that some words, like "to table" have the opposite meanings in British and American English. When Americans say "table" the British say "shelve" and when the British say "table" the Americans say "put forward" or "propose.")
So "British English" just refers to speaking English with one of many British accents, I'm guessing in this case approximately London English that is not Cockney.
1) You should fully Americanize all the spellings. It is American English people want.
2) This would be far too boring for me to stick with if it is things like "I went to the cinema yesterday". A coherent story, or even better, a domain-specific lesson, would be a much more engaging way to teach a language. I was able to learn when it was "talk to my driver about the security situation and drive plan", but never cared enough for casual conversation. I am usually happy to talk to people who speak horrible English about things I care about, which presumably for the hn audience is tech, startups, etc., but not about sports (cricket!?). If you could do a vertical-specific sayafter.me it would be awesome.
I agree with te_chris, it is not only arrogant but incorrect to make that assumption. I don't actually mind using 'color', but you live in a small world if you think that only a certain accent is relevant.
(I prefer South African English, personally, but I think the commercial market is strongly in favor of American English.)
Same thing in China and Thailand; I don't know about other countries.
More anecdata for you.
"No, your anecdotal evidence is wrong."
"My anecdotal evidence is right."
(A totally fluent UK/IE/etc English speaker probably is fine in any context, but I don't think someone learning from a webapp is starting from that point. If you are going to be a moderate speaker, it is better to be moderate in the target accent.)
I think one advantage I have over this product is, SayAfter.me stores ever growing stats of word/phrase pronunciation attempts from all users.
Later on it can build custom exercises depending on the background of a particular user.
I will be able to say "You are Spanish, other Spanish users had the most trouble with these words. So here is an exercise that contains problematic English words for native Spanish speakers".
This thing wouldn't have been possible without Google really.
She doesn't mispronounce the words she practised anymore.
Also I got positive feedback from the early beta users (other students & English learners) so I assumed it was useful in its initial state.
But yes, I will add additional information with benefits and comparison with other products.
Another nice hack for practising speech is calling free phone support of various companies/products and tell them about your problem with their product/service. I learnt this from an Italian guy.
You mean "my pronunciation." English is pretty weird about when we do and do not use the plural form and this trips up a lot of people :) I wouldn't mention it, but given that you're here for the specific purpose of improving your English, I thought you'd want to know.
Speaking of which, mass nouns always trip non-native speakers up, especially in IT. Do NOT use the following words: datas, informations, softwares. Data is already the plural of the Latin word datum and none of those words are used by native speakers.
I would add 'codes' to that use as it is often misused, but there are a lot of times when using it is correct, so you'll just have to study up on the rules for mass nouns sometime.
It's not the most natural thing to say but it can still fly:
"Merriam-Webster provides a free online dictionary, thesaurus, audio pronunciations, [...]"
in a way that 'softwares' and 'advices' would not.
I would add 'codes' to that use as it is often misused
To make this even more confusing, in some contexts, 'codes' is used as a plural in ways you never would for general purpose programming, even though it refers to the same thing. Physics (nuclear, in particular) simulations are a case in point:
Which is, of course, quite harmful to someone who wants to improve. And yes, I know of no simple answer as to when to use 'code' vs. 'codes', but I will say that I think that 'source codes' is probably always wrong. At least, I am unable to find a sentence using that which sounds right, though it might be possible for it to be correct if you were talking about something other than the source code of a computer program.
My filter might be a bit more lenient because I imagine an implied 'my pronunciations of some particular words" after it. I still wouldn't say it or write it but it doesn't poke me in the eye as much, especially in the middle of otherwise idiomatic-sounding casual English. So context counts as well - as do the tastes of the person reading your stuff. Someone might think you an illiterate clod for a misused "it's" or "your" or "whomever".
Anyway, I think we're actually in violent agreement, my only point is the nuances are tough and not subject to easy rules.
Oh, and you were saying
At least, I am unable to find a sentence using that which sounds right, though it might be possible for it to be correct if you were talking about something other than the source code of a computer program.
Then again, I seem to recall some observation about the evolution of language where the less-used irregular words change more frequently, and 'source code' is not something people outside of technology discuss. But that's probably why the usage, which makes them seem like outsiders, sounds bad to me.
Will definitely check this one out.
Btw: Yes, my email address ends in .de - your registration form seems to think I mistyped and suggests that I change that to .net. Intentional, even for valid TLDs?
If this keeps happening often maybe I should use my own domain list.
Thanks for the feedback.
You probably should do it as soon as possible :
> The included default top level domains are com, net, org, info, edu, gov, co.uk, and mil.
Which means ... it excludes all regional non english TLDs. Pretty much a problem for a service that targets non english native speakers. :)
By the way, I am an english non native speaker that writes a lot of english through the internet but never ever speak english. This service is just perfect for me, thank you.
I will also be adding lots of new content soon, so if you are interested in a particular theme for words/phrases ("flowers", "dog breeds", "Flirting 101" etc) just let me know. I will give them higher priority.
I have no special theme right in mind, I usually speak english only when I travel, so it's mainly general purpose language. This could be as well asking for menu in restaurant, discussing tech with other developers or debating politics with foreign friends.
On the other hand, the idea of isolating words that are especially often wrongly pronounced for a given native language you mentioned in an other comment sounds really awesome : I'm less interested in getting a specific word right than getting right specific sounds used in a lot of words (but I guess this does not help you to select your first words, sorry about that).
Actually this is the long term goal. If I ever have the chance to work on SayAfter.me full time, I will map phonetic sounds to words, and find out which nationalities have trouble with which sounds.
More info here: http://en.wikipedia.org/wiki/IPA_number#Letters