Languages are identified on this site by three-letter ISO 639-3 codes.

These codes provide a consistent, international way to identify individual languages. Since many languages are known by several names, these codes helps to identify the specific languages listed on my site, to avoid confusion between languages with similar names, and to avoid duplicate listings (such as one language being listed under several different names). The use of ISO 639-3 also helps to distinguish between languages in different parts of the world that share the same name, such as Kayan (which may refer to Kayan of Burma [pdu], Kayan of Papua New Guinea [kct], Kayan of Malaysia [kys], Kayan of Indonesia [xkn], etc.), and Lele (which may refer to Lele of Chad [lln], Lele of the Democratic Republic of the Congo [lel], Lele of Guinea [llc], Lele of Papua New Guinea [lle], or even Lyélé of Burkina Faso [lee], which is also sometimes called "Lele").  Although many languages share the same name, each individual language has its own three-letter ISO 639-3 code.  

More information about ISO 639-3 codes is available from Ethnologue's page explaining three-letter codes for identifying languages and from The Linguist List's answer to the question What is an ISO 639-3 code?

ISO 639-3 has a code for each language; The Linguist List's Multitree takes this a step further by assigning an additional three letters for dialects of each language. 

For example, the code for Swiss German is [gsw]; within this language there are a variety of dialects such as the dialects of Basel [gsw-bas], Bern [gsw-ber], Zurich [gsw-zur], and Alsatian [gsw-als].  This allows a specific dialect of a language to be identified more precisely.

Every language has variations; these differences may be minor or very large. 

In linguistic terms, generally two varieties of speech are considered "different languages" if their speakers cannot understand each other; if they are similar enough so that a speaker of one variety can understand the other without difficulty, or read the same literature, they are usually considered "dialects" of the same language. 

In some parts of the world, certain languages are referred to as "dialects" for social or political reasons, not linguistic ones. 

The official language of a country might be called a "language" while the other languages spoken in that country are referred to as "dialects", usually implying that the "dialect" is less developed, less educated, or less important than the "language".  This is not a linguistic classification.  In linguistic terms, it would be more correct to describe them as "minority languages" (if they are significantly different from other languages spoken in the area) or "non-standard dialects of the majority language" (if the differences are sufficiently minor). 

The way people commonly classify languages does not always match their linguistic classification. 

For some languages, the most common ways of describing the languages do not match their classification in linguistic terms.  Usually, this means that several individual languages are commonly classified as a single language.  This can mean that two people who are said to speak the same so-called "language" (usually actually a macrolanguage) may actually not be able to communicate with each other. A Chinese-speaker who speaks Mandarin, for example, is unlikely to be able to understand a speaker of Yue Cantonese, even though both speak "Chinese".  Additionally, the language that is commonly spoken in a given region is not necessarily the language that is taught in schools or used for reading and writing; therefore, many speakers of regional languages are also able to read and write the more widely-used language (whether or not that written language reflects the spoken language they use). This fact can cause some linguistic confusion as well, since some people may incorrectly assume that different people who are able to read the same material must also speak a common, single language.

Example 1: Chinese. Chinese is a macrolanguage [zho], not an individual language. It includes thirteen different languages, including Gan Chinese [gan], Hakka Chinese [hak], Huizhou Chinese [czh], Jinyu Chinese [cjy], Mandarin Chinese [cmn], Min Bei Chinese [mnp], Min Dong Chinese [cdo], Min Nan Chinese [nan], Min Zhong Chinese [czo], Pu-Xian Chinese [cpx], Wu Chinese [wuu], Xiang Chinese [hsn], and Yue Chinese [yue].  These languages are different enough that speakers of one variety of Chinese usually cannot understand speakers of another variety.  Most use the same writing system (with minor variations), however, so speakers of different varieties of Chinese are often able to read the same written material.  Even for writing, however, there are regional differences, as well as two written forms: Simplified Han Script [hans], used mostly in the People's Republic of China and Singapore, and Traditional Chinese [hant], used mostly in Taiwan, Hong Kong, and Macau.

Example 2: Arabic. The name "Arabic" can refer either to a language (Standard Arabic [arb]), or to a macrolanguage (Arabic [ara]). The macrolanguage includes more than thirty different spoken languages, including Algerian Spoken Arabic  [arq], Baharna Spoken Arabic [abv], Chadian Spoken Arabic [shu], Egyptian Spoken Arabic [arz], Gulf Spoken Arabic [afb], Hadrami Spoken Arabic [ayh], Hijazi Spoken Arabic [acw], Libyan Spoken Arabic [ayl], Mesopotamian Spoken Arabic [acm], Moroccan Spoken Arabic [ary], Najdi Spoken Arabic [ars], Omani Spoken Arabic [acx], Sa’idi Spoken Arabic [aec], Sanaani Spoken Arabic [ayn], Shihhi Spoken Arabic [ssh], South Levantine Spoken Arabic [ajp], Sudanese Spoken Arabic [apd], Ta’izzi-Adeni Spoken Arabic [acq], and Tunisian Spoken Arabic [aeb].  As indicated by the word "Spoken" in the names used by Ethnologue for these languages, these are the languages commonly spoken (but not usually written) in different regions.  Standard Arabic [arb] is the language that is primarily used for reading and writing, so educated speakers in most of these regions are also able to read, write, and speak Standard Arabic [arb] in addition to speaking the regional variety. 

Example 3: German. The German spoken in Switzerland is often considered a regional dialect of Standard German [deu], but it is actually different enough from Standard German to be linguistically considered a separate language, Swiss German [gsw]  Ethnologue (17th edition) indicates that 93% of German speakers in Switzerland speak Swiss German [gsw], although Standard German [deu] is used as a language of instruction in schools.  For a comparison of greetings in varieties of German and related languages, click here.

Even in linguistic terms, the difference between a "language" and a "dialect" is not always clearly defined. 

Speakers of one language might be able to understand another to some extent, but not the other way around.  For example, many speakers of Estonian can understand Finnish to some extent, but Finnish speakers generally have a more difficult time understanding Estonian; likewise, it is usually easier for speakers of Dutch to understand Afrikaans, than it is for Afrikaans speakers to understand Dutch (in both of these cases, the differences are significant enough that they are classified as separate languages).  Some aspects of a language may be very similar to another language or dialect, while other aspects might be significantly different.

Example 1: Mixtec. The language of Mexico commonly referred to as "Mixtec" is actually a large group of multiple languages. They are not considered dialects because speakers of one variety are not normally able to understand speakers of another.  Ethnologue lists 52 Mixtec languages; INALI lists more than 80.  The Mixtec greeting is Taon in Magdalena Peñasco [xtm], Tanikun in Alcozauca [xta], Tanikuu in Metlatónoc [mxv], Tániku in Silacayoapan [mks], Takuni in Mixtepec [mix], Takuni vei in Southeastern Nochixtlán [mxy], Á nákaaní in Tezoatlán [mxb], Lamanuni  in Apoala-Apasco [mip], Nanova'ani in Chazumba [xtb], Koba'ani in Southern Puebla [mit], Kobaana in Peñoles [mil], and Ku va'a in Ataláhuca [mib].  Some similarities between these languages are evident; the Mixtec languages form a continuum, meaning that languages change gradually from one region to the next. Nevertheless, the differences are significant enough that each variety is considered to be a separate language, not simply a dialect of a single language.  For a list of greetings in Mixtec and other Otomanguean languages, click here.

Example 2: Yokuts. The Yokuts language [yok] of central California is often classified as one language with many dialects.  However, the various dialects are different enough that they are considered by many people to be different languages. The Yokuts greeting, literally meaning "where?", is Hel in Tulamni, Hille in Chukchansi, Hielu in Chawchila, Xiw'eu in Toltichi, Hiyeuk in Tachi and Wakichi, Hiyuk in Choynok and Yawelmani, Hide in Yawdanchi, and Hite in Wikchamni.  Instead of classifying all of these varieties as a single language, it might make sense to classify them into groups such as Buena Vista Yokuts (Tulamni), northern Yokuts (Chukchansi, Chawchila, and Toltichi), southern Yokuts (Tachi, Wakichi, Choynok, and Yawelmani), and Tule-Kaweah (Yawdanchi and Wikchamni); or to classify each variety as a separate language.  However, since the ISO 639-3 standard classifies them as a single language, I've done the same on my website.

Example 3: ISO 639-3 codes for Akan and Norwegian. There are a few cases in which even the ISO 639-3 standard leaves the distinction between languages and dialects ambiguous.  Akan [aka] has its own language code, but so do its two main subdivisions: Fanti [fat] and Twi [twi]. This leaves unresolved the question of whether to consider Fanti and Twi "dialects of Akan" or "two separate languages" (on my website I've chosen to consider them separate languages).  A similar issue exists with Norwegian [nor], which has two forms, Bokmål [nob] and Nynorsk [nno], both of which have their own ISO 639-3 codes in addition to the inclusive code [nor] for Norwegian.  As with Fanti and Twi, I've included both on my website.

The classification of languages and dialects often changes.

Although ISO 639-3 codes provide a consistent, international way to identify individual languages, they also change as well, as linguists understand the relationships between languages and dialects in new ways.  I try to keep up with these changes on my website, but may miss some. I'm also aware that not all linguists or speakers of the languages in question agree with the classifications used by ISO 639-3. It's inevitable that there will be some errors (which I'd love to know about; please email me) and updates as new information becomes available or as linguists decide on a different way to classify languages for a variety of reasons.

Although I have used greetings as examples above, greetings and similar phrases often can be among the worst ways to judge how similar one language or dialect is to another.

To accurately determine the relationships between languages and dialects, linguists look at a large number of basic words, as well as many other aspects such grammar, pronunciation, written forms, contact with speakers of other languages and other historical factors, etc.  Different groups often use different ways of greeting even if their languages are otherwise very similar.  Others borrow greetings form other languages or use fossilized, archaic word forms in greetings even though those words are not used in other contexts. Greetings are one very small part of a language, and often do not even provide a typical example; greetings and polite phrases may contain sounds or linguistic structures not used elsewhere in the language.

Example 1: Portuguese and Chuwabu. The fact that speakers of both Portuguese [por] of Portugal and Chuwabu [chw] of Mozambique greet by saying Bom dia does not mean that their languages are similar; it just indicates that Chuwabu speakers borrowed the Portuguese greeting. Other aspects of the two languages are very different from each other. 

Example 2: Lezgi, Sindhi, Tausug, and Haroi. Likewise, speakers of Lezgi [lez] of Russia, Sindhi [snd] of Pakistan, Tausug [tsg] of the Philippines, and Haroi [hro] of Vietnam all greeting by saying Salam; this does not mean that they speak the same language or even related languages, but rather that they all borrowed the greeting from Arabic. 

Example 3: Swiss German. The reverse is also true: the fact that a speaker in one part of Bavaria says Grüß Gott while someone from Südbaden says Salli and someone in Zurich says Gu-ëte Morgë does not mean that they speak significantly different languages, but simply that they are saying greetings that have different meanings - "greet God", "hi", and "good morning" - in closely related dialects of Swiss German [gsw]. 

Example 4: Ga and Burmese.  In addition to borrowed phrases, there are also linguistic coincidences: the Ga [gaa] greeting Miiŋa bo of Ghana and the Burmese [mya] greeting Mingala ba of Myanmar look and sound similar (note that the letters ŋ and ng represent different ways to spell the same sound), but they are completely unrelated; Miiŋa bo means "I am greeting you" in Ga while Mingala ba, etymologically from the Pali language [pli] of Myanmar and India, means "auspiciousness to you".

Resources

Hammarström, Harald, Robert Forkel, Martin Haspelmath, and Sebastian Bank. 2015. Glottolog 2.4. Leipzig: Max Planck Institute for Evolutionary Anthropology. Online: http://glottolog.org/

INALI. 2010. Catálogo de las Lenguas Indígenas Nacionales. México D.F.: Instituto Nacional de Lenguas Indígenas. Online: http://www.inali.gob.mx/clin-inali/

Lewis, M. Paul, Gary F. Simons, and Charles D. Fennig (eds.). 2015. Ethnologue: Languages of the World, Eighteenth edition. Dallas, Texas: SIL International. Online: http://www.ethnologue.com

Lewis, M. Paul.  2013. "So What's in a Name?" Ethnoblog. Dallas, Texas: SIL International. Online: https://www.ethnologue.com/ethnoblog/paul-lewis/so-whats-name

Lewis, M. Paul, Gary F. Simons, and Charles D. Fennig (eds.). 2015. "The Problem of Language Identification." Ethnologue: Languages of the World, Seventeenth edition. Dallas, Texas: SIL International. Online: https://www.ethnologue.com/about/problem-language-identification

Lewis, M. Paul, Gary F. Simons, and Charles D. Fennig (eds.). 2014. "Three-Letter Codes for Identifying Languages." Ethnologue: Languages of the World, Seventeenth edition. Dallas, Texas: SIL International. Online: http://www.ethnologue.com/codes/default.asp

Linguist List. 2014. Multitree: A digital library of language relationships. Bloomington, IN: Department of Linguistics, The LINGUIST List, Indiana University. Online: http://multitree.org/

Linguist List.  2014. "What is an ISO 639-3 code?" Multitree: A digital library of language relationships. Bloomington, IN: Department of Linguistics, The LINGUIST List, Indiana University. Online: http://linguistlist.org/forms/langs/find-a-language-or-family.cfm#iso-code

SIL International. 2015. "Dialects". ISO 639-3: Scope of Denotation for Language Identifiers. Dallas, Texas: SIL International. Online: http://www-01.sil.org/iso639-3/scope.asp#Dialects

SIL International. 2015. "Macrolanguages." ISO 639-3: Scope of Denotation for Language Identifiers. Dallas, Texas: SIL International. Online: http://www-01.sil.org/iso639-3/scope.asp#M

Simons, Gary, and Steven Bird. 2015. "OLAC Language Extension." OLAC Reommendations. Philadelphia, Pennsylvania: University of Pennsylvania.  Online: http://www.language-archives.org/REC/language.html