The idiom is the traditional answer indeed, although there are different versions of it. You may find that "a language has an army and a navy, a dialect does not" or "a language has an army or a navy, a dialect does not".

This maxim is from Europe, the 18th or 19th century. This was before an air force was a critical component.

But that fact is important. As nation-states were developing in the "we have guns now" age, there was a great call for standardization of language. This would make it easier to communicate throughout the nation, keep track of the citizens, educate them, and generally make a linguistic boundary to reinforce a political boundary.

With the cataclysmic European-lead wars of the first half of the 20th century, many nation-states weakened, and regional languages and dialects were reemerging from the official blacklist. Languages like Welsh, Basque, Sardinian, Sorbian, and others became once again speakable. Official recognition varied and continues to vary. In some cases, official status has been granted to former suppressed languages (like Catalan), and some have engaged in armed struggle as a compliment or core component of seeking recognition (Basque, Irish).

The status of minority languages in the Americas, Asia, and Africa varies. In some cases, you have many official languages (South Africa, India). In others, you have no official language and an effort made in certain areas to create a linguistic environment that allows speakers of multiple local languages to access important services (United States; my medical system papers come with a note in about 20 languages explaining how to get a translated copy or interpreter). Some have what outsiders (and many insiders) consider multiple languages, but are officially considered one language (China). In others, native languages still thrive, but the dominant language is that of the Indo-European invaders and those who rely on native languages are considered uncultured (various areas of central and south America, but the remnants of native language speakers in the US and Canada are treated in a very mixed way).

The modern humanist perspective is to treat forms of speech as languages when their users consider them languages. Ethnologue takes this kind of approach. Other classification systems would consider many languages in Ethnologue as dialects.

As for Costa Rica and Iceland and their armies, they are rare and new exceptions.

Iceland was settled in the 9th century and has maintained its language since settlement. But Iceland is small and isolated with nearly no immigration or emigration. There are less than 300,000 people there. There are records of everyone who ever lived there. Iceland also was a colony of Denmark for most of its history. And it was extremely poor, to the level of having famines, with no hope of assistance. Its current wealth is very recent, a 20th century phenomenon.

And part of that is because Iceland does have a de facto army, navy, and airforce, that of the United States. Iceland was used as an air base and naval base by the US in WW2 and it continues to exist deep into NATO territory.

Central America has been considered the playground of the US military for 200 years. Costa Rica is a demilitarized zone in an area where the US military has been active for a very long time. Costa Rica is home to a significant proportion of American citizens, especially the retired. In that sense as well, Costa Rica has a military, it just belongs to the US. Both of its neighbors have been the subject of US military intervention, and Panama was created by the US fomenting an independence war to free it from Colombia so that the Panama Canal could be made.

But Costa Rica doesn't have an indigenous language that is prevalent, it uses Spanish like much of the Americas.

The question of what is a language and what is a dialect itself is a modern question. Until there were sufficient enough people who could write and it became permissible to use something other than Latin, it was mostly just a question of whether you could understand what the other person was saying. If you couldn't, could you find someone who could translate between the two of you? Yes? Good enough.

Thanks Maria...I did not know that people have tried to pin it down to an actual percentage of overlap. Do you happen to know a good citation for that? And what about dialects?

And thank you very much Larisa for the idiom. That is actually more along the lines of relevance to my paper, which will not delve too far into this topic because its not something I know about and I donʻt want to open a can of worms (where worms are critiques) by making like I know anything about linguistics.

The idiom is the traditional answer indeed, although there are different versions of it. You may find that "a language has an army and a navy, a dialect does not" or "a language has an army or a navy, a dialect does not".

This maxim is from Europe, the 18th or 19th century. This was before an air force was a critical component.

But that fact is important. As nation-states were developing in the "we have guns now" age, there was a great call for standardization of language. This would make it easier to communicate throughout the nation, keep track of the citizens, educate them, and generally make a linguistic boundary to reinforce a political boundary.

With the cataclysmic European-lead wars of the first half of the 20th century, many nation-states weakened, and regional languages and dialects were reemerging from the official blacklist. Languages like Welsh, Basque, Sardinian, Sorbian, and others became once again speakable. Official recognition varied and continues to vary. In some cases, official status has been granted to former suppressed languages (like Catalan), and some have engaged in armed struggle as a compliment or core component of seeking recognition (Basque, Irish).

The status of minority languages in the Americas, Asia, and Africa varies. In some cases, you have many official languages (South Africa, India). In others, you have no official language and an effort made in certain areas to create a linguistic environment that allows speakers of multiple local languages to access important services (United States; my medical system papers come with a note in about 20 languages explaining how to get a translated copy or interpreter). Some have what outsiders (and many insiders) consider multiple languages, but are officially considered one language (China). In others, native languages still thrive, but the dominant language is that of the Indo-European invaders and those who rely on native languages are considered uncultured (various areas of central and south America, but the remnants of native language speakers in the US and Canada are treated in a very mixed way).

The modern humanist perspective is to treat forms of speech as languages when their users consider them languages. Ethnologue takes this kind of approach. Other classification systems would consider many languages in Ethnologue as dialects.

As for Costa Rica and Iceland and their armies, they are rare and new exceptions.

Iceland was settled in the 9th century and has maintained its language since settlement. But Iceland is small and isolated with nearly no immigration or emigration. There are less than 300,000 people there. There are records of everyone who ever lived there. Iceland also was a colony of Denmark for most of its history. And it was extremely poor, to the level of having famines, with no hope of assistance. Its current wealth is very recent, a 20th century phenomenon.

And part of that is because Iceland does have a de facto army, navy, and airforce, that of the United States. Iceland was used as an air base and naval base by the US in WW2 and it continues to exist deep into NATO territory.

Central America has been considered the playground of the US military for 200 years. Costa Rica is a demilitarized zone in an area where the US military has been active for a very long time. Costa Rica is home to a significant proportion of American citizens, especially the retired. In that sense as well, Costa Rica has a military, it just belongs to the US. Both of its neighbors have been the subject of US military intervention, and Panama was created by the US fomenting an independence war to free it from Colombia so that the Panama Canal could be made.

But Costa Rica doesn't have an indigenous language that is prevalent, it uses Spanish like much of the Americas.

The question of what is a language and what is a dialect itself is a modern question. Until there were sufficient enough people who could write and it became permissible to use something other than Latin, it was mostly just a question of whether you could understand what the other person was saying. If you couldn't, could you find someone who could translate between the two of you? Yes? Good enough.

Hi everyone! I see the initial question as a timelessly interesting and relevant one; and all the better, posed in a rather innocent guise. Then Larisa Fialkova came, I feel, to mark the anything-but-innocent socio-political side to it.

The way I read it, there is indeed a lot more politics in this than is often acknowledged. To bring but a few well-known cases, dialects in China will not be signed off as languages of their own, so long as the corresponding government favours national unity with a vengeance. Now, do Cantonese and Mandarin speakers understand a word of the other language?

Far from the People's Republic, in Spain there is an interesting Catalan triangle. Catalan has been an official language of the country since Franco died, and democracy took over. Catalonians welcomed this as an only fair release of the heavy thumb that had repressed them for 40 years. And now, a few decades later, some Valencians -- who a speak a dialect of Catalan -- want their dialect to be signed off as a language of its own. Well, do you think they have much support from Catalonians in that enterprise? Nay. Can you find any similarity with the case of China?

In sum, what linguistic regulations have in common worldwide is politics, and various interests therein. The sooner we acknowledge this, the less energy will be wasted in many a province.

U zijn in Tilburg! Vlaams en nederlands, twee of één taaltje?

The Netherlands is a major center for the study of language - experimental, historical, multilingualism, sociolinguistics, anything really. A culture very open to linguistic exploration, whether on a scientific level or an everyday level.

But if you want to see a really complicated analysis of linguistic variation, take a look at an article called "Quantitative Social Dialectology: Explaining Linguistic Variation Geographically and Socially."

As the abstract says they are examining pronunciation distances between standard Dutch and 424 Dutch dialects. Yes, 424 dialects in 42,000 km2.

(the article is free to download)

Yes and no, Maria.

The Netherlands has a very complicated history. During the time of the Roman Empire, the Netherlands was entirely above sea level with a large lake in the middle. South of the lake was Roman, north was Frisian. Over the next 1000 years global warming and glacial rebound resulted in about half of that are being underwater. For most of that time, they were nominally part of the Holy Roman Empire, but no one cared because they were a damp forest with some cow farmers.

Things changed. Flanders, especially Brabant (half in Netherlands, half in Belgium) became the wealthiest part of western Europe in the 1000-1400ad time, based on making textiles from imported wool, silk, cotton, etc. Around this time as well, ships got better and Constantinople was conquered by the Ottomans.

Now that forest in Holland (which means "forest land" was valuable for ship-building, and Venice could no longer import fancy goods from the east. A growing population also required protein, in the form of salted fish, which Holland also had plenty of.

This gave them money to begin dredging the land from the sea. It also attracted the attention of the dominant empire of the day, Spain, and most of the Netherlands and what is now Belgium became the property of Habsburg Spain. But parts remained in the HRE, and the Frisians remained independent.

Protestantism then made a mess of things, with the north (the Netherlands) becoming mostly Protestant and the south (now Belgium) mostly Catholic. This didn't suit Spain or the Netherlands, and an 80 year war for independence was waged and won.

At this point, the Netherlands stopped being a bunch of independent feudal lands and islands, and became a nation, dominated by Holland.

But the Catholic south was not happy with Protestant domination and won independence in 1830, giving us Belgium.

So the small area of the Netherlands has a large amount of linguistic diversity.

First, much of it was a chain of islands or peninsulas for about 1000 years, and linguistic drift occurs rapidly in an island environment.

Second, the parts that weren't islands were influenced by different neighboring areas. In the north and east, the language is closer to low (Saxon) German than the Holland dialect. In the southeast, the influence is from middle Rhine German.

Third, the Frisians remained unwilling participants in the country, and Frisian is essentially a different language from both Dutch and German.

Fourth, Holland and Frisia had more interaction with the English than the rest of the country, while Zeeland and Brabant had more contact with France and Wallonnia.

And finally, we have a population that has been particularly literate and interested in linguistics for a few hundred years. There was much less official pressure from the central government on provincial languages, and more focus on examining language differences than in most of the world.

So, yes, 424 dialects is crazy. But a split into Holland, Zeeland, Brabant, Limburg, Twente, Groeningen, and Frisian is realistic. And even within that, most of Zeeland, Holland, and large parts of Groeningen were islands for hundreds of years.

If you look at the map below, you can see just how much of the country was essentially an archipelago even as late as 1579:

http://upload.wikimedia.org/wikipedia/commons/4/4c/Tachtigjarigeoorlog-1579.png

Thank you very much James for the contribution! I found that paper fascinating, and very relevant to my region of study, which is Polynesian Island groups and the cultural (not necessarily lingual, but language is certainly part of it) variation both within and between the island groups. So the Netherlands divergence and convergence based on their temporary status as an archipelago was very interesting. I had no idea they were so segregated so recently in history. I also really appreciated that the paper discussed a bit how some of the population and influence imbalance seemed to drive some of the differences. That is exactly what I am interested in the moment, some of the drivers of cultural differences, particularly when starting with common culture. So thank you for the excellent contribution!

Maria, I agree with you about the dialect/location issue - there aren't 424 dialects spoken in the Netherlands, and they are taking samples from 424 locations. The surprising thing to me is that the article does specifically refer to them as 424 dialects, in the second sentence of the abstract and elsewhere.

I'm surprised that the reviewers allowed that term to be used. I've read the article, and from many (or most) of the locations they have a single person as the source of pronunciation.

But here's another map:

http://www.wetenschap24.nl/nieuws/artikelen/2011/mei/Prehistorische-landkaarten.html

Click down a little and it shows the Netherlands in 800ad (the year Charlemagne was crowned emperor in Rome).

Peter M. Gardner University of Missouri-Columbia (retired)

Morris Swadesh's renowned, systematic research on this goes back over 60 years. But, his two basic articles that provide detail on his proposed techniques are:

"Toward greater accuracy in lexicostatistical dating," International Journal of American Linguistics 1955, v. 21:121-37.

"Some limitations of diffusional change in vocabulary," American Anthropologist 1956, v. 58:301-6.

I used his approach in determining the relationships of India's Dravidian Languages in 1980, but my article in Indian Linguistics contains too many typographical errors to be a good guide. The printers even got Swadesh's name wrong!

Thank you Peter...I will look into Swadeshʻs work.

One good feature of Swadesh's lexicostatistical (LS, also known as glottochronology) approach is that, as a result of decades of work on the historical linguistics of hundreds of languages, we are able to make systematic, research based "trees" of the dialects and languages of whole language phyla. I should have mentioned that in my earlier response. LS also entails well-established, rigorous techniques. It has been found that in any given language family, there tends to be a certain rate of change over time (usually approximately 81% of basic vocabulary is retained per millennium). Using the basic vocabulary and establishing the retention rate are both critical. Also, LS findings are routinely and systematically cross-checked against all available historical and archaeological findings. Historical linguists who do such research work as scientists in the strictest sense of that term. Note that you cannot do this LS work if you are only concerned with a few languages or dialects. You would simply lack the necessary body of data.

Why aren't people today conversant with all this? Many of the suggestions you have received from others simply wouldn't begin to provide the standards achieved using Swadesh's techniques. It may be of interest to you (as a New Zealander) that such research was done on Malayo-Polynesian languages in the 1950s.

Ah yes....I have come across this a bit in my research of the Pacific, by Yen and others looking at the dialectic distance of the Polynesian languages. It is a robust and rigorous study. I think the work in Polynesian could be redone now since the linguistics of each island group have been much better documented than when the previous work was done, which only utilized a limited subset of common words. I agree with you that this seems the only truly rigorous work that I have seen around linguistics. I did not know that it stemmed from Swadash...itʻs always good to know the academic genealogy of these studies. In answer yo your question I donʻt know why people are not more conversant with all of this. Although not a linguist and barely an anthropologist I still find this fascinating, especially when you consider the rapid loss of languages occurring these days.

A language can IMO best be considered to be (a group of) mutually understandable speeches & dialects. Different countries, however, often call their own dialect a language (Scandinavian, Serbo-Kroatian, Szecho-Slovak), although Dutch & Belgian Nederlands (together 6 dialect groups) are the same (written) language (Dutch). Letzeburgian is usu.considered to be a different language from German (Deutsch), although the Luxemburg dialect doesn't differ much from the neighbouring German dialect. Dialects can understand each other with a bit difficulty, e.g. Afrikaans & Dutch are usu.considered to be different languages, although they were written the same until the 1920s IIRC. I can understand written Afrikaans with little effort, but have serious difficulties in following 2 Afrikaners talking to each other. Dutch & Frisian are clearly different languages. Portuguese (European & Brazilian), Spanish (Castilian & Hispano-American), Catalan, Provinçal, French, Italian (incl. Sardinian?) & Rumenian can be considered different languages (although Catalan & Provinçal are often considered to be "only" dialects). The 3 Walloon dialects (have difficulties understanding each other) are considered to be French dialects. Languages sometimes change very rapidly (wars, invasions, migrations), sometimes very slowly (church Latin). I read that Icelandic children can read texts about 800 years old, but in Dutch this is totally impossible. IOW, the difference between language & dialect is political, subjective & gradual.