Really, how did we decide that the word 'spoon' had to mean a spoon and not tree? Or perhaps we didn't decide anything but as it happens, King Charlemagne built a bridge over the river Rhine. The bridge was built in ten years and was a famous bridge at that time. However, soon after he died, the bridge was destroyed by a conflagration so completely that a contemporary account said, "nought oon spoone" could be seen floating on water. Spoon then meant a splinter or a section of wood and it evolved semantically from Old English into our modern usage.
If you have been a student in any Pakistani school you would have heard at least one of your teachers poke fun at English phonology when they smugly ask why the words 'but' and 'put' are pronounced differently as in /bʌt/ and /pʊt/ respectively. While they may not know of a weird convention which says that a vowel preceded by a labial (that is, said with lips like in /p/, /f/, /b/) consonant will be pronounced rounded. However, the word 'but' still sticks out as an exception to this rule and isn't pronounced rounded. But this exception is mainly based on locals accents as a Northern Britisher pronounces both but and put in a similar fashion (rounded). (If you have ever heard Tan France say the word 'but' you would know what I mean). Whilst the teachers in our schools may not know of this weird rule, haply, they also don't know the fact that put and but are two different phonemes and can have different pronunciations nonetheless.
Language is by far the best thing to have happened to human civilization. There are a myriad of theories to describe the evolution of languages. Noam Chomsky's big idea is that language is pre-wired in some way in our brains. The only problem with this theory is that we don't know which part of the genes holds the code for languages (or even if they do in the first place). Plenty of other theorists have spoken of gesture-first theory. That is to say: we first started using hand-gestures to communicate, and only later did it occur that we moved from gestures to speech. The neural areas that control gestures lie close to those that control the mouth – and in infancy too, gestures precede language. Again the refutation catches up with the theory when we realize that using gestures is calorically nonviable and humans, in general, are lazy. Simply put, gestures are too much effort. Also, pantomime is a loose art: how does one pantomime for, let's say, being depressed?
There is yet another theory which propounds that language evolved culturally some 40,000-50,000 years ago with the onset of the Acheulean Age when better bifacial stone tools were carved and average brain size increased. Again, the art of tool creation can be learnt (and passed on) by imitation and doesn't necessarily need language. Guy Deutscher in his book The Unfolding of Languages has tried to deconstruct the complex nature of today's languages. Much of the reasoning that follows in this essay draws on his book (and another beautiful book by David Crystal called A Little Book of Language) to outfox the language labyrinth.
Treading on the terra incognita, and given that our ancestors knew basic action words (run, eat, climb) pointing words (this, that) and a few nouns, we can use the paradigm of 'Economy, Expressiveness, Analogy' to arrive at the colorful world of modern languages. We will only tackle 'Economy' as a driving force for the invention of new words in this essay. Just to give a sneak peek, Analogy is used to construct words by looking at the pattern of present words and constructing new ones on the same pattern. For instance, let's look at the word 'grotty' which happens to be an adjective as it is formed on the basic template of adjectives that is Noun+y (as in luck+y, chubb+y, class+y), the only problem is that when this word was invented in the 1970s, no noun for the adjective 'grotty' existed. Surely there was the word 'grotesque' but it's just another adjective. So to make up for this discrepancy the noun for this word was also invented after-the-fact. Hence, we had the noun 'grot' and its adjective 'grotty' made out of thin air. This is how analogy works.
To explain expressiveness, take another word 'irony' which comes from Latin īrōnīa, from Old Greek εἰρωνεία, and from εἴρων meaning "one who feigns ignorance." Juxtapose 'irony' with 'one who feigns ignorance', and I think we can conclude that we are better served by the word 'irony'. Now, we shall return to inventing new words by using the principle of least effort. In Italian, the word caldo means 'hot', while in English hot means 'hot (or warm)' but the roots for both words come from Proto-Indo-European words kel (warm) and kai (burn) respectively. As it happens, the sound K of Kai was weakened to Ch and then to H in order to save effort. Producing the sound [K] consumes more effort than the sound [H]. Consequently, we went from Kai to Ch to H. (Kai>Chai>Hai-ta>Hot). Other sounds like [P] and [T] were also weakened to [F] and [Th] to save effort in the Germanic languages of the Proto-Indo-European language family.
This weakening of the consonant was driven solely to save effort, where the guiding principle was to 'pronounce as little as possible.' Words can sometimes change shape or meaning without creating something new or lessening the effort. For instance, the meaning of the word 'resent' changed from 'taking something with a good feeling' to its complete opposite. The accusative-case for the verb 'dive' is changing from 'dived' to 'dove' in a fashion similar to drive>drove. (Analogy!)
Lastly, we talk a little about how our national language made it as our national language. The vernacular in the Mughal courts of the Indian subcontinent was Persian, so when the Mughals mingled with locals speaking Hindustani, they gave birth to a pidgin language. A pidgin language is born when two disparate languages intersect with each other. The distinct languages gave birth to a language that lacked grammar and native speakers, yet it was halfway between the two languages it stemmed from. This Persianized Hindustani thus became Urdu, once it got native speakers - by which we mean the people who learnt Urdu as their first language. Urdu-Hindi both have the same registers (stylistic variations .i.e jargon, colloquism, slang etc) and perhaps that is why so much of Bollywood makes sense in Pakistan.