In Hindi grammar, उपसर्ग (Upsarg) and प्रत्यय (Pratyay) are the two primary tools of शब्द-निर्माण — word formation. Together, they explain how a single root word (मूल शब्द) branches into dozens of related words.
Upsarg literally means "what is placed before" — it attaches to the beginning of a root word and modifies or intensifies its meaning. Think of it like a qualifier you prepend: the root word "गति" (motion) becomes "प्रगति" (progress), "अवगति" (decline), or "दुर्गति" (bad state) simply by swapping the prefix. The root meaning does not disappear; the prefix steers it.
Pratyay works the opposite way — it attaches to the end of a root word, typically to change its grammatical category or express a quality. "मानव" (human being) becomes "मानवता" (humanity) by adding the suffix "ता". The verb root "पछता" becomes the noun "पछतावा" (regret) with the suffix "आवा".
An everyday analogy: think of the root word as a plain paratha. Upsarg is what you add on the left side of the plate — a pickle, a chutney — that changes the character of the meal. Pratyay is what you put on the right — butter, curd — which changes the texture or form. Same base, completely different dish depending on what you add and where.
The exam does not test you on obscure words. It tests whether you can correctly identify which part of a given word is the upsarg or pratyay — a task that trips candidates who memorise word lists but never understand the mechanics of how the attachment actually works. The key skill is knowing where the root ends and the affix begins.
There is one more wrinkle: Hindi draws upsargs from three distinct sources — Sanskrit (तत्सम), Hindi-native (तद्भव), and Persian/Arabic (विदेशज). Each group has its own set, and the exam sometimes tests whether you know which tradition a prefix belongs to.
An upsarg by definition cannot stand alone — it has no independent meaning as a free word. Its job is purely morphological: attach, modify, done.
These are the most tested group. There are 22 standard Sanskrit upsargs. The ones that appear most in UP Police exams:
| Upsarg | Core meaning | Example | |--------|-------------|---------| | प्र | forward, onward | प्रगति, प्रबल | | प्रति | against, each, back | प्रतिकूल, प्रत्येक | | अनु | following, after | अनुकरण, अनुशासन | | अव | down, away from | अवगत, अवनति | | सम् | complete, with | सम्पूर्ण, संयोग | | परि | around, completely | परिक्रमा, परिचय | | वि | separation, special | विशेष, विघटन | | सु | good, well | सुगम, सुपुत्र | | दुर् / दुस् | bad, difficult | दुर्गम, दुस्साहस | | अ / अन् | negation | अज्ञान, अनादर | | अधि | over, upon | अधिकार, अध्यक्ष | | उप | near, secondary | उपकार, उपनेता | | नि | down, into, without | निगम, नियम | | निर् / निस् | without, out of | निर्भय, निस्संदेह | | परा | away, back | पराजय, पराक्रम | | अभि | toward, intensive | अभिमान, अभिनय | | अति | beyond, excess | अतिरिक्त, अत्यधिक | | उत् / उद् | up, above | उत्पत्ति, उद्यम | | आ | toward, from | आगमन, आकर्षण |
Critical identification rule for 'प्रति': When you see a word starting with "प्रत्य-", do not be fooled by the visual form. Sandhi is happening: प्रति + एक = प्रत्येक. The upsarg is still "प्रति", not "प्रत्य". The exam specifically exploits this — see the PYQ on "प्रत्येक" below.
A smaller group, but tested. Key ones:
| Upsarg | Meaning | Example | |--------|---------|---------| | अ / अन् | negation | अधूरा, अनपढ़ | | अध | half | अधपका, अधमरा | | उन | one less | उनतीस, उनसठ | | भर | full | भरपूर, भरसक | | सु | good | सुडौल, सुघड़ |
These appear in Urdu-origin words common in Hindi usage:
| Upsarg | Meaning | Example | |--------|---------|---------| | बद | bad | बदनाम, बदतमीज़ | | बर | on, upon | बरदाश्त, बरकरार | | ब | with, by | बाकलम, बनाम | | ना | negation | नाकाम, नालायक | | हम | co-, fellow | हमदर्द, हमराह | | बे | without | बेकार, बेशर्म | | कम | less | कमज़ोर, कमअक्ल | | खुश | happy, good | खुशनुमा, खुशबू |
Note on "ब" vs "बद" vs "बर": These look similar but are distinct. "बाकलम" uses "ब" (meaning "with/by pen"). "बदनाम" uses "बद" (bad name). "बरदाश्त" uses "बर". Confusing these three is one of the most common exam errors.
Pratyay attach at the end and are categorised into two major types:
These attach to verb roots (धातु) and form nouns or adjectives.
| Pratyay | Example | |---------|---------| | आवा | पछतावा (पछता + आवा), चढ़ावा | | आवट | सजावट, बनावट | | आवत | कहावत (कह + आवत), सुहावत | | आई | लड़ाई, पढ़ाई | | अन | लगन, चलन | | इया | रसिया, डाकिया |
These attach to nouns, adjectives, or other non-verb bases.
| Pratyay | What it creates | Example | |---------|----------------|---------| | ता | abstract noun (quality) | मानवता, सुंदरता | | पन | abstract noun (state) | बचपन, लड़कपन | | त्व | abstract noun (Sanskrit) | मनुष्यत्व, देवत्व | | इक | adjective from noun | मानवीय → actually "ईय"; सामाजिक | | ई | adjective or noun | दिल्ली, पहाड़ी | | वान / मान | possessive adjective | गुणवान, बुद्धिमान |
Identifying pratyay correctly: Strip the suffix and check if a recognisable root remains. "कहावत" → remove "आवत" → "कह" (verb root: to say). That confirms "आवत" is the pratyay. If you try removing only "वत", you get "कहा" — that looks like a valid word but is not the standard root here, so "वत" is not the pratyay.
When a word begins with "प्रत्य-", "प्रत्यु-", or "प्रत्यव-", immediately recognise that sandhi has merged "प्रति" with a following vowel. Pattern: प्रति + vowel = प्रत्य + vowel.
Examples: प्रति + एक = प्रत्येक | प्रति + उत्तर = प्रत्युत्तर | प्रति + अय = प्रत्यय
So the upsarg is always "प्रति", never "प्रत्य". Standard approach: guess the shape you see (30s of confusion). This pattern: 5 seconds, zero doubt.
Three pratyays that all create abstract nouns denoting a quality or state: ता, पन, त्व.
Quick rule: "ता" and "त्व" usually attach to Sanskrit/tatsam bases (मानव → मानवता, देव → देवत्व). "पन" attaches to Hindi-native bases (बच्चा → बचपन, लड़का → लड़कपन).
When an exam option says "मानवता" and another says "मानवत्व", both are technically formed correctly — but "मानवता" is far more common in standard usage and almost always the exam answer.
Standard approach: guess randomly among abstract noun options (50% error rate). This pattern: identify the base language → correct answer in 8 seconds.
Three Persian prefixes that trip candidates: बद (bad/evil), बर (on/upon), ब (with/by).
Test: does the word carry a negative moral meaning? → बद (बदनाम, बदतमीज़). Does it carry a sense of "tolerating" or "maintaining a position"? → बर (बरदाश्त, बरकरार). Does it carry a sense of "by means of" or instrument? → ब (बाकलम = by pen, बनाम = in the name of).
When you see four options mixing बद/बर/ब/बे, eliminate by asking: is the meaning "bad", "upon", "by means of", or "without"? This cuts 4 options to 1 in under 10 seconds vs the 40-second habit of trying each combination.
Strip the suspected suffix from the end and check if the remainder is a recognisable Hindi verb root or noun. If yes, the stripped part is the pratyay. If no, you stripped too much or too little.
Example: "पछतावा" — try removing "वा" → "पछता" (this is a valid verb: to regret). But "वा" alone is not a standard pratyay. Now try "आवा" → "पछत" — not a standard root. So strip "आवा" → "पछता" — valid root, valid pratyay.
This test takes 4 steps vs randomly matching against a memorised list. Reduces wrong choices from 3 to 0 in under 15 seconds on a 4-option question.
Two suffixes that look similar: आवा (action result noun: पछतावा, चढ़ावा) and आवत (proverbial/habitual noun: कहावत).
Key: "आवत" almost exclusively appears in words that have a proverbial or habitual character — कहावत (saying/proverb), सुहावत. "आवा" appears in words that name the result of an emotional or physical action.
When the exam question has both as options, ask: is this word a proverb/saying-type noun? → आवत. Is it the result of an action/feeling? → आवा. This distinction solves the कहावत vs पछतावा type questions in 10 seconds.
Look at the given word and apply this decision sequence in the exam hall:
Step 1 — Identify the question type. Is it asking for the upsarg (prefix) or pratyay (suffix)? Do not skip this — they look similar in options and reading fast leads to answering the wrong question.
Step 2 — Locate the root. Mentally strip away what you suspect is the prefix or suffix. Does a recognisable Hindi/Sanskrit word remain? If yes, you have found the root.
Step 3 — Check for sandhi masking. If the word starts with प्रत्य-, प्र्यु-, etc., apply the PRATI-Vision rule before anything else.
Step 4 — Use source-language clues. If the word has Urdu/Persian feel (contains ब, بد sounds, etc.), look at Persian prefix options first.
Step 5 — Eliminate by meaning. If two options look similar (e.g., "आवा" vs "वा"), apply the Root-Strip Test to determine which stripping leaves a valid root.
Time budget: This question type should not take more than 25-30 seconds. If you are stuck after Step 4, mark your best guess and move on — do not let a single grammar question eat into your GK and Reasoning time.
Why this question: The exam exploits the sandhi transformation of "प्रति" into "प्रत्य" to trick candidates who read only the visual beginning of the word.
Solving path: The word is "प्रत्येक". Apply the PRATI-Vision rule: "प्रत्य" = "प्रति" + vowel (here, "एक" starts with "ए"). So the upsarg is "प्रति", not "प्रत्य". Option C is correct. Discard "इक" (that's a taddhit pratyay, not an upsarg), "प" (a Hindi-native upsarg, not present here), and "प्रत्य" (not a real upsarg — it's a sandhi form).
Why this question: "पछतावा" is a standard example of the कृत् प्रत्यय "आवा" — the exam presents "वा" and "आ" as distractors that result from stripping too much.
Solving path: Apply Root-Strip Test. Remove "आवा" → root = "पछता" (to regret — a valid verb). This is clean. Remove "वा" → root = "पछता" still works, but "वा" is not a standard Hindi pratyay. Remove "आ" → root = "पछताव" — not a standard root. Therefore "आवा" is the pratyay. Option B.
Why this question: Tests whether you know that "ता" is the abstract noun suffix that converts "मानव" into "मानवता" (humanity).
Solving path: The question asks what "मानव" becomes with the suffix "ता". Apply: मानव + ता = मानवता. The other options: "मानवीकरण" uses suffix "ईकरण", "मानवीय" uses "ईय", "मनुष्य" is an unrelated synonym. Option D is correct.
Why this question: "प्रतिकूल" is a frequently tested word because it cleanly demonstrates "प्रति" in action, and "प्र" is a close look-alike distractor.
Solving path: Strip "प्रति" → root = "कूल" (bank/shore; here used in the sense of direction/current). Strip "प्र" → "तिकूल" — not a word. "परा" gives you words like "पराजय", not "प्रतिकूल". "परी" is not a Sanskrit upsarg. Option A is correct.
Why this question: "कहावत" vs "आवत" vs "वत" — this question directly tests the AAVAT suffix knowledge and whether candidates confuse it with "वत".
Solving path: Apply Root-Strip Test. Remove "आवत" → root = "कह" (verb: to say — valid). Remove "वत" → "कहा" — looks valid but "वत" is not a standard pratyay in Hindi word formation. Remove "हावत" → "क" — not a root. "कह" as root confirms the pratyay is "आवत". Option C is correct.
Why this question: Tests the Persian prefix "ब" specifically — distinguishing it from the more commonly known "बद" and "बर".
Solving path: "बाकलम" = ब + अकलम (meaning "by pen" — Urdu/Persian origin). The prefix "ब" means "by/with". "बदनाम" uses "बद" (bad). "बरदाश्त" uses "बर" (upon/endure). "बदौलत" uses "बद" variant meaning "by virtue of" — actually "ब" + "दौलत", but the exam key confirms "बाकलम" as the "ब" upsarg word. Option A is correct.
Confusing "प्र" with "प्रति": These are two separate upsargs. "प्रगति" uses "प्र". "प्रतिकूल" and "प्रत्येक" use "प्रति". The visual similarity catches candidates who do not check the root that remains after stripping.
Treating "प्रत्य" as an upsarg: "प्रत्य" is never an upsarg in itself — it is always "प्रति" that has undergone sandhi with a following vowel. Every time you see "प्रत्य-", think "प्रति + vowel".
Stripping too little from pratyay words: In "पछतावा", removing only "वा" leaves "पछता" which looks right but "वा" is not a valid pratyay. The full "आवा" must be identified.
Mixing up abstract noun suffixes "ता", "पन", "त्व": All three make abstract nouns. Candidates often mark "त्व" for "मानवता"-type words. Check the specific word in the question — the exam will only have one correct form.
Confusing बद/बर/ब in Persian prefix questions: These look similar. Spend 5 seconds mentally asking: "bad quality?→बद", "tolerate/maintain?→बर", "by means of?→ब". Do not guess among these three.
Identifying the root incorrectly when both prefix and suffix are present: Some words have both an upsarg and a pratyay (e.g., "अनुशासन"). When the question asks for only one, strip the correct end. Stripping from the wrong end gives you a nonsense root and a wrong answer.