Thoughts on: What is the Chinese language?

The Economist recently shared two opposing views of Chinese linguistics.  I have included them in bold with my comments.

In brief, Chinese traditionalists believe:

1) Chinese is one language with dialects.
I would argue this is a highly political statement to make.  As part of its political rhetoric, China states that it is one people comprised of fifty some minorities.  Having everyone, from different ways of life and different cultures, speaking different dialects from the same language, works to reinforce that notion.  One might disagree, but how many times have pro-Tibet groups argued for Tibetan independence by citing the fact that Tibetans do not speak Chinese or that Chinese language is forced upon them?

2) Chinese is best written in the character-based hanzi system.
How does one define best? See below.

3) All Chinese read and share the same writing system, despite speaking in different ways.
All Chinese is a bit extreme. Mongolians, Manchus, Tibetans (just to name the most obvious examples) all have their own writing systems distinct from Chinese characters.  Most now can read and write in Chinese characters due to the education system, but this was by no means a natural progression.


Western linguists tend to respond:

1) Chinese is not a language but a family; the "dialects" are not dialects but languages.
Western linguists define dialects as part of the same language if they are intercomprehensible.  If they are not, then they are distinct languages, not dialects. If a Mandarin speaking Beijinger listened to a Cantonese speaker, he or she would not understand what was being said.  If one accepts this definition, then one must accept this conclusion. 

With that said, many regional dialects/languages are, however, intercomprehensible.  Chinese in the same province or same region can often understand each other's dialects even though they cannot speak them.

2) Hanzi-based writing is unnecessarily difficult; the characters do not represent "ideas" but "morphemes" (small and combinable units of meaning, like the morphemes of any language). Pinyin (the standard Roman system) could just as easily be used for Chinese. Puns, wordplay and etymology might be sacrificed, but ease of use would be enhanced.
Chinese characters are in part composed of morphemes (items that represent meaning), but there are a lot of characters with phonetic indicators (meaning they indicate pronunciation).  Regardless of how one looks at the linguistic analysis, what is meant by best represented?  Arguably, one could communicate with just pinyin, although there would undoubtedly be the occasional confusion that would not have arisen if characters were used. 

However, this is not the question at hand.  Is the best writing system to use one that solely reflects pronunciation?  One could question, is the Roman alphabet the best writing system for English?  In English, there are countless of words for which the writing does not reflect pronunciation. Languages such as French and German have entered the English language, and while the written word has retained its original form, the pronunciation is distinctly English.  One could very well argue that we could rid ourselves of this by using a phonetic system such as IPA.  Aside from being more work than it's worth, there is a history that is lost in the transition. Most Chinese would say the same about their own language. 

In addition, pinyin itself is not free from problems.  Some sounds and letters are not represented consistently.  For example in yan and yang, the vowel a makes different sounds in each.  With chu and qu, the written form of the consonants change while the vowels remain constant, while for the actual pronunciation the opposite is true.

3) Modern hanzi writing is basically Mandarin with the old characters in a form modified by the People's Republic. Everyone else (Cantonese speakers, say) must either write Mandarin or significantly alter the system to write their own "Chinese".
I have heard Chinese people speaking formally in dialect, and the structure is similar to how it would be in Mandarin.  However, is this the natural way in which dialects are expressed in formal discourse or is it the result of Mandarin language education?  When dialect speakers write colloquial sentences, they use different Chinese characters to express themselves than they would in Mandarin, and a non-speaker of the dialect would not necessarily understand.  I am certainly not the expert on this, though. Any thoughts?