I should add: yes, while the four Vedas (Rigveda, Yajurveda, Samaveda, and Atharvaveda) were passed down for about three thousand years with a high degree of fidelity using a complex system of oral memorization techniques (in fact, until fairly recently it was even considered taboo to write the Vedas down*), and while we can assume that the Vedas as they are are handed down reasonably accurately, that doesn’t mean that the transmission was 100% perfect, or that there are absolutely no variants or recensions of the same texts (there are) or changes to the texts along the way.
For an example of ‘alteration’ (not always necessarily intentional), you have for example sound changes.
The Vedas, for example, has mainly come down to us in two main versions: ‘continuous’ recitation (saṃhitā-pātha), the ‘normal’ version that students of the Vedas first learn and the one that is commonly chanted, and the deritative ‘word by word’ recitation (pada-pātha), where the whole text is broken up into its basic components. Let’s say for example, the second verse of the very first hymn of the Rigveda:
“Worthy is Agni [the god of fire] to be praised by living as by ancient seers; he shall bring hitherward the gods.”
Saṃhitā: Aghniḥ pūrvebhir ṛṣibhir īḍyo nūtanair uta / sa devām̐ eha vakṣati
Pada: agniḥ | pūrvebhiḥ | ṛṣi-bhiḥ | īḍyaḥ | nūtanaiḥ | uta | saḥ | devān | ā | iha | vakṣati
You might notice the linguistic phenomenon called *[sandhi]("[I) (saṃdhi) in the saṃhitā version. Sandhi refers to the “combination” of two sounds that sit next to each other. Let’s say, for example the two words synthesis and symbol: both words contain the same prefix syn-, but before the letters ‘b’, ‘m’ and ‘p’ the ‘n’ of syn- changes to ‘m’ due to sandhi (hence ‘symbol’, ‘symmetry’, and ‘sympathy’ versus ‘synthesis’ or ‘synopsis’).
To go back to the example, in the phrase pūrvebhir ṛṣibhir īḍyo nūtanair uta, ‘worthy to be praised by ancient sages (rishis) and present ones’ you could notice sandhi working: -ḥ for example changes to -r, -aḥ to -o.
One way of ensuring that the hymns - the mantras - were passed down accurately was to apply rules of pronunciation and sandhi onto the hymns, which however were not always accurate, and sometimes, even obscured the original meaning. (Which is not necessarily a drawback: in Indian thought, the words themselves, not so much the meaning of the words, are what was important and seen to carry power.)
One of the famous Vedic mantras is the so-called Gayatri mantra. The word gāyatrī refers to a particular Vedic poetic meter: three lines with eight syllables each, which is what the mantra (originally) was. The mantra was originally an invocation to the solar deity Savitr, one of many Vedic gods that have fallen by the wayside in later Hinduism. (So the interesting thing is, Savitr as a god is no longer really worshiped, but the mantra bearing his name continues to have an important position in Hinduism not unlike the Our Father in Christianity.) The mantra goes:
Tat Savitur vareṇyaṃ
bhargo devasya dhīmahi
dhíyo yo naḥ pracodayāt
“May we attain that excellent glory of Savitr the god:
So may he stimulate our prayers.”
Now here’s the thing. The mantra as handed down isn’t completely metrical: the first line has just seven syllables (tat-sa-vi-tur-va-re-ṇyaṃ). It turns out that originally, the word vareṇya would have been pronounced as vareṇiya, thus yielding eight syllables: tat savitur vareṇiyaṃ. The later simplification of -niya to -nya obscured the original poetic meter.
- For an interesting comparison, ancient Persians/Iranians - who were descended from the same ancestors as the Indo-Aryan peoples of India, the Indo-Iranian nomads - once had the same sentiment towards the sacred hymns of Zoroastrianism (which are in Avestan, a language - or rather two different dialects of the same language - closely related to Sanskrit, the sacred language of the Indo-Aryans; in fact, Avestan is closer to the ancient Sanskrit of the Vedas than Vedic Sanskrit is to later Sanskrit).
Zoroastrian priests at first considered writing to be the domain of ‘barbarians’ (three thousand years ago, only their Semitic neighbors really wrote down stuff) and insisted at first on passing down their knowledge orally, even after the Persians had developed their own writing system. Compared to Indians however, the Persians adopted the idea writing down a sacred text fairly quickly; by the 3rd-4th century, Zoroastrian sacred texts were already being written down. However, at the same time, they made sure that the text represented the sounds of Avestan (as it was passed down in their time) accurately.*