Scarcity is the first thing every collector reaches for when justifying a price. "It's an edition of 89 — of course it's expensive." But the Shepard Fairey secondary market, read through 48 of his prints where we hold both a verified edition size and a documented sale history, tells a more disciplined story. Small editions do command higher unit prices. They also trade far less often, price-discover poorly, and can strand a seller who needs liquidity. Larger editions give up headline price but win on turnover — and in a market with no third-party authenticator, turnover is what actually protects your exit.
This is Post 2 in our data-driven Shepard Fairey series. The numbers below come from a single source: our cleaned comp set of 79 Fairey prints, 48 of which now carry a confirmed edition size and release year. Every median price, sale count, edition figure, and dated transaction is drawn from that set. Where the data is thin or silent, we say so plainly rather than fill the gap with a story. The goal is not to sell you on scarcity or to debunk it, but to measure exactly what it buys and what it costs.
The scarcity thesis, stated plainly
The classic argument runs like this: fewer copies means more buyers chasing each one, which means a higher clearing price and stronger appreciation over time. It is intuitive, it is partly true, and it is also the single most over-applied idea in the print market. The problem is that "scarcity" is treated as a one-way lever — as if a smaller edition can only help. In reality, edition size is a trade-off. Shrinking the pool lifts the price a buyer will pay on any given day, but it also shrinks the number of days on which any sale happens at all. A print you cannot sell for eighteen months at $600 is, for many purposes, worth less than a print you can sell next week at $250.
Across our 48 edition-bearing Fairey prints, the confirmed edition sizes span from 80 (The Greed Depression, 2023) to 700 (the two Tunnel Vision Version 2 variants, 2018). The catalog-wide edition range is even wider — our structural data records a floor of 40 and a ceiling of 2,100 across all catalogued works — but within the comp set, where we can actually observe transactions, 80 to 700 is the working band. That is an important distinction. The 40-to-2,100 span describes what Fairey has printed; the 80-to-700 span describes what actually trades often enough to price. The scarcity thesis has to be tested against the second range, because a rumored edition of 40 with no comps is not a market, it is a curiosity.
To test it, we grouped the 48 prints into four edition bands and computed, for each band, the median of the per-print median prices and the average number of tracked sales per print. Every figure in the table below is arithmetic on the 48-record set only.
Table A — Fairey prints bucketed by edition-size band
| Edition band | Prints in band | Median price (median of medians) | Avg. tracked sales per print | Total tracked sales |
|---|---|---|---|---|
| ≤150 | 10 | $231 | 23.9 | 239 |
| 151–350 | 10 | $288 | 53.1 | 531 |
| 351–550 | 21 | $200 | 76.7 | 1,611 |
| 551+ | 7 | $190 | 59.4 | 416 |
Read the two right-hand columns together and the shape of the market appears. Liquidity climbs steadily as editions grow — the average print jumps from roughly 24 tracked sales in the tightest band to 77 in the 351–550 band, more than tripling. The 351–550 band alone accounts for 1,611 of the set's tracked sales, because that is where Fairey actually prints most of his benchmark runs of 400, 450, and 500. Meanwhile median price does not climb with scarcity in a clean line. It peaks in the middle (151–350) and is lowest, not highest, at the top of the edition range. The smallest band sits at $231 — respectable, but below the 151–350 band's $288.
That non-monotonic result is the first crack in a naive scarcity model. If scarcity alone drove price, the ≤150 band would top the table by a wide margin. It does not, because edition size is only one of several inputs — subject, format, year, and cultural weight all pull on the number, sometimes harder than scarcity does. What edition size does do reliably is govern liquidity, and that shows up unmistakably in the sales columns. The tightest editions are the hardest to trade, full stop.
Small editions lift unit price — the correlations
Two correlations computed across all 48 prints make the trade-off explicit. The correlation between edition size and per-print median price is -0.21 — weakly negative, meaning larger editions are associated with somewhat lower unit prices, not higher ones. The correlation between edition size and tracked sale count is +0.31 — positive, meaning larger editions trade more often. Neither is a strong linear relationship (Fairey's market is too heterogeneous for that, and 48 points is a modest sample), but the signs are the story: scarcity nudges price up and liquidity down, exactly as the trade-off predicts. A market where scarcity were a free upgrade would show a positive price-to-edition correlation approaching zero from below; instead we get a clear negative sign on price and a clear positive sign on volume.
A cleaner way to see it is to split the 48 prints at their median edition size of 400 and compare the two halves head to head.
Table B — Small editions vs. large editions, split at the median (edition 400)
| Group | Prints | Median price | Avg. tracked sales per print |
|---|---|---|---|
| Smaller editions (≤400) | 26 | $254 | 43.3 |
| Larger editions (>400) | 22 | $189 | 75.9 |
The split is decisive. Prints at or below an edition of 400 carry a median price of $254; prints above 400 carry $189. That is a 34% unit-price premium for the smaller half — computed as $254 divided by $189. So the scarcity premium is real: on a like-for-like day, the smaller edition clears higher. But look at the sales column. The larger half averages 75.9 tracked sales per print against 43.3 for the smaller half — roughly 75% more transactions. Scarcity bought a third more price and paid for it with nearly double the illiquidity.
This is the central finding of the post, and it deserves to be stated without hedging: in Fairey's market, scarcity lifts the price you can ask but crushes the frequency at which anyone answers. For a collector who buys and holds forever, the premium is a gift — they will never need the liquidity they are giving up. For an investor who may need to exit inside a defined window, the illiquidity is a tax, and it is a tax that does not show up in a headline median. A $254 median on a print that sells four times a year is a different asset than a $254 median on a print that sells four times a month, even though the two look identical on a valuation sheet.
The anchor cases: where scarcity pays
Abstract correlations are easy to wave away, so consider the specific prints that carry the thesis. These are the small-edition premium payers in our set, and each one earns its price a slightly different way.
THESE SUNSETS ARE TO DIE FOR (LARGE FORMAT) is the cleanest example of scarcity working. Released in 2023 at an edition of just 89 — the second-smallest edition in the entire comp set — it carries a median of $650, an average of $659, and a recent average of $655 across 43 tracked sales. That is a print holding firmly in the mid-hundreds while most large-format contemporaries in the set trade below $300. The large physical format compounds the small edition: fewer copies, each a wall-dominating object. The price history confirms the level rather than a fluke — eBay results of $700 (Jan 6, 2026) and $716 (Jan 13, 2026) sit right on top of the median, with a softer WorthPoint print of $447 (Apr 19, 2026) reminding us that even a strong small edition has a spread. Notice, too, that with 43 sales it is not illiquid by the standards of a sub-100 edition — it prints small but sells steadily, which is the rare combination collectors pay up for.
Peace Goddess (2006, edition of 300) is the highest median in the entire 48-print set at $700, with an average of $828 and a recent average of $849 across 31 sales — one of the few prints in the set where recent trades sit above the long-run median, a sign of live appreciation. Its edition of 300 is not tiny, but pair a mid-size edition with an 18-year vintage and a marquee image and you get a price no recent release in the set matches. The dated tape shows the range in full: a $1,368 eBay result (Nov 5, 2024), a $980 sale (Jan 6, 2026), and a $700 WorthPoint print (Jan 7, 2026). Even the down-comps in its history — a $275 WorthPoint result back in 2019 — are the exception, not the trend. Scarcity here is doing its work in concert with age, and the combination is the most powerful pairing in the set.
Lenin Record (2005, edition of 300) tells a subtler version of the same story and a cautionary one. Its median is $525 and its long-run average is a striking $900 — but its recent average has cooled to $410 across just 16 tracked sales. The gap between a $900 lifetime average and a $410 recent average is exactly the kind of signal a thin-liquidity print produces. Its dated history swings violently: a $125 eBay result (Jul 3, 2025) sits a few months before a $700 eBay sale (Jan 6, 2026), with a $375 WorthPoint print (Mar 30, 2025) in between. With only 16 sales on the tape, a handful of early auction-house results can inflate the average while the current market clears anywhere across a wide band. Small-and-vintage can carry a rich price, but 16 data points is a fragile foundation. We flag this openly — the median is the more trustworthy number here, and even that rests on a slim sample.
Basquiat Canvas (2010, edition of 450) is worth a mention because it shows a large-ish edition still commanding a premium price — a median of $475 on 19 sales — through format and subject rather than scarcity. It is a canvas, not a paper print, and it references one of the most valuable names in twentieth-century art. Edition 450 would ordinarily place it in the moderate-price workhorse band; instead it ranks fourth by median in the whole set. Format and subject overrode edition. Hold that thought — it recurs.
The counter-case: where volume beats scarcity
Now the print that should not exist if scarcity were the whole story. A CHAMPION OF JUSTICE (Ruth Bader Ginsburg), released in 2021 at an edition of 500 — a large run by Fairey standards — carries a median of $460, an average of $553, and a recent average of $706, all on a remarkable 227 tracked sales, the deepest liquidity of any print in the set.
By the naive scarcity model this is a contradiction: a large edition should mean a soft price. Instead RBG delivers the fifth-highest median in the 48-print set and more than five times the sales volume of the median print. The explanation is not scarcity — it is demand. The RBG image is culturally loaded, tied to a moment and a figure with mass appeal well beyond Fairey collectors, and it was released into a broad, motivated audience. The large edition did not suppress the price; the demand simply overwhelmed the supply. The dated tape underlines the strength: a $950 eBay result (Feb 17, 2026) and repeated results near $700 in late listings, against a launch-window print of $550 in March 2021. A recent average of $706 sitting well above the $460 median, on a 227-sale sample, is about as trustworthy an appreciation signal as this market produces.
RBG is the exception that clarifies the rule. Scarcity is a lever, but it is subordinate to demand. A small edition of a print nobody wants is not valuable; it is merely rare. A large edition of a print everybody wants is a liquid blue-chip. When you evaluate a Fairey, ask about the demand first and the edition second — because demand is the variable that can override every other input, and edition size is not.
Table C — High price / low liquidity vs. lower price / high liquidity
| Profile | Edition | Median | Tracked sales | |
|---|---|---|---|---|
| High price, thin liquidity (scarcity-led) | Peace Goddess (2006) | 300 | $700 | 31 |
| THESE SUNSETS (LARGE FORMAT) (2023) | 89 | $650 | 43 | |
| Lenin Record (2005) | 300 | $525 | 16 | |
| Lower price, deep liquidity (volume-led) | Kurt Cobain – Endless Nameless (2021) | 650 | $200 | 167 |
| Universal Dignity (2022) | 600 | $200 | 149 | |
| Power and Glory Letterpress (2016) | 450 | $180 | 177 | |
| Rare hybrid (demand overrides edition) | A CHAMPION OF JUSTICE / RBG (2021) | 500 | $460 | 227 |
The two profiles are almost mirror images. The scarcity-led prints cluster at $525–$700 on 16–43 sales apiece. The volume-led prints cluster near $180–$200 but on 149–177 sales — four to ten times the trade frequency. Then RBG breaks the frame entirely, delivering a scarcity-tier price on volume-tier liquidity, precisely because demand, not edition size, was the binding constraint. If you were forced to pick the single most attractive financial profile in the table, it is not the $700 median on 31 sales — it is the $460 median on 227 sales, because it pairs a strong price with an exit you can actually use.
The volume-led names reward a closer look, because they are where most collectors will actually transact. Kurt Cobain – Endless Nameless (2021, edition 650) carries a $200 median on 167 sales and a tight price band — its long-run average of $208 and recent average of $204 sit almost on top of the median, with dated eBay results clustered from $125 to $280. That tightness is the gift of liquidity: with 167 comps, the price is discovered to within a narrow range, and a buyer or seller can act with confidence. Universal Dignity (2022, edition 600) tells the same story — a $200 median, a $203 average, a $209 recent average across 149 sales, its dated history hugging the $195–$240 range. Power and Glory Letterpress (2016, edition 450) rounds it out at a $180 median on 177 sales. None of these will make a collector rich, but none will trap one either. They are the fixed-income of the Fairey market: modest coupon, high liquidity, low drama. The scarcity chasers may scoff, but a portfolio built partly on names like these is a portfolio you can actually rebalance.
Set that liquidity against DALAI LAMA COMPASSION (2022, edition 500), which shows what happens when a large edition carries a strong subject but the market has not fully settled. Its $400 median sits well above the volume-led cluster on 118 sales, yet its recent average has slipped to $314 and its dated tape swings from a $665 WorthPoint result in 2018 down to $206 in mid-2026. Even with triple-digit liquidity, a print in the middle of re-pricing can whipsaw — a reminder that liquidity tightens discovery but does not freeze the level. Demand still moves the number; liquidity just tells you how reliably you can read it.
Scarcity does not equal appreciation
A separate claim often rides shotgun with the scarcity thesis: that small editions appreciate faster. Our set lets us test it crudely by asking how many prints in each half show a recent average above their long-run median — a rough proxy for a print trading up. The answer is a near-tie. In the smaller-edition half (≤400), 16 of 26 prints show recent averages above their medians. In the larger half (>400), 14 of 22 do. Proportionally that is 62% versus 64% — effectively identical. Scarcity, on this measure, confers no appreciation edge at all.
What actually separates the appreciators from the laggards is not edition size but the same demand factor that explained RBG. The strongest upward moves in the set — RBG's recent average of $706 over a $460 median, Celebration Day's late eBay results of $840 and $935 against a $305 median, Peace Goddess's $849 recent average over a $700 median — belong to prints with strong subjects and, in two of three cases, editions of 300–500 that are anything but tiny. Celebration Day (2012, edition 300, 89 sales) is instructive: its dated history runs from a $132 WorthPoint print in 2014 to $840 and $935 eBay results in early 2026, then back to $155 in April 2026. That is a print that has genuinely re-rated over a decade, but the re-rating rode on the image and the vintage, not on a scarce edition — 300 is squarely mid-size.
The lesson for anyone treating Fairey prints as an appreciating asset: do not pay a scarcity premium expecting it to double as an appreciation premium. They are different things. The appreciation, when it comes, comes from demand catching up to an image over time — and it happens to large editions and small editions alike.
Vintage cuts deeper than edition
If edition size is a weak predictor of price, what is a strong one? In this set, age. Splitting the 48 prints into vintage (2015 and earlier) and contemporary (2016 and later) produces a gap far larger than anything edition size generates.
Table D — Vintage vs. contemporary Fairey prints
| Cohort | Prints | Median edition | Median price | Avg. tracked sales |
|---|---|---|---|---|
| Vintage (2015 & earlier) | 11 | 400 | $345 | 45.2 |
| Contemporary (2016 & later) | 37 | 400 | $200 | 62.2 |
The two cohorts share an identical median edition of 400 — so edition size is held roughly constant — yet the vintage cohort's median price of $345 is 72% higher than the contemporary cohort's $200 (computed as $345 divided by $200). That is more than double the scarcity premium we measured in Table B, and it appears even though the two groups are matched on edition. In other words, being ten-plus years old does more for a Fairey's price than halving its edition does. The vintage cohort also trades a bit less often (45 sales versus 62), consistent with older prints being tightly held, but the price gap dwarfs the liquidity gap.
This reframes the scarcity conversation entirely. When a 2005 Lenin Record or a 2006 Peace Goddess commands a rich price, it is tempting to credit the edition of 300. The vintage cross-cut suggests the age is doing at least as much of the work — and probably more. A collector chasing value should weight release year heavily and edition size lightly, which is close to the opposite of how most scarcity arguments are framed. We develop this in the dedicated blue-chip tier analysis; here it is enough to note that edition size ranks behind both demand and vintage as a price driver.
Format is the other silent variable
Edition size also says nothing about format, and format moves price hard. The four canvas prints in our set — Basquiat Canvas ($475, ed 450), Keith Haring Canvas ($394, ed 450), John & Yoko Canvas ($370, ed 450), and Tom Petty: An American Treasure Canvas ($388, ed 275) — all sit in the upper half of the price distribution despite editions of 275–450 that would otherwise place them mid-pack. Three of the four share an edition of exactly 450, the most common workhorse size in the set, yet they price like scarce objects. The canvas format, not the edition, is carrying that premium.
The same logic runs through THESE SUNSETS (LARGE FORMAT), where the parenthetical is doing real financial work: large format plus small edition is why it clears $650. Strip the format out and an edition of 89 alone would not guarantee that level. The takeaway is consistent with everything above — edition size is one input among several, and when it is pitted against format, subject, or vintage, it usually loses. Read an edition number as a modifier on value, never as its source.
Why liquidity is the number investors underweight
A median price tells you what a print is worth on the day it sells. A sale count tells you how often that day arrives. Collectors fixate on the first number and quietly ignore the second, and that is a mistake with real financial consequences in a market like Fairey's — where, as our authentication work stresses throughout this series, there is no third-party COA program. Value rests entirely on the artist's signature, the edition numbering, provenance, and condition. There is no certificate to lean on, so the market's collective willingness to transact is the authentication signal. A print with 177 documented sales has been vetted 177 times by buyers putting real money down. A print with 16 has been vetted 16 times, and you are relying far more heavily on your own eye.
Liquidity does three concrete things for a holder. First, it tightens price discovery: with 100-plus comps, the median is a tight, trustworthy estimate; with 16, it is a wide guess wrapped in a number. Second, it shortens your exit: a print that trades weekly can be sold when you choose, not when a rare buyer happens to materialize. Third, it dampens variance: deep markets absorb a motivated seller without a price collapse, while thin markets gap down on a single anxious listing. The Lenin Record spread between a $900 lifetime average and a $410 recent average is that variance made visible — and the $125-to-$700 swing in its dated history is the same fact from another angle.
There is a fourth, quieter benefit: liquidity is a proxy for demand, and demand is what appreciates. The deepest-liquidity prints in our set — RBG (227), Power and Glory Letterpress (177), MAKE ART NOT WAR (176), Kurt Cobain (167) — are prints the market keeps coming back to. High turnover is not merely a convenience at exit; it is evidence that the image has durable pull. A print nobody re-lists is telling you something.
It is worth putting a dollar frame on the illiquidity cost, because "harder to sell" is abstract until it hits your timeline. Consider two hypothetical sellers, both needing to raise cash in ninety days. The first owns a print from the ≤150 band averaging 24 sales across the multi-year window in our data — call it a sale every seven to eight weeks in a healthy stretch, and far less often in a slow one. Ninety days may see one comparable transaction, or none, which means the seller either waits or discounts to force a buyer. The second owns a 351–550 workhorse averaging 77 sales — several a month — and can list into a live, self-refreshing market that clears near the median without a fire-sale cut. Same nominal value, radically different cash-conversion profile. The scarcity premium the first seller paid at purchase is, in that ninety-day window, working against them. This is the practical face of the +0.31 liquidity correlation: it is not a statistic, it is the difference between selling on your schedule and selling on the market's.
None of this means small editions are bad buys. It means the scarcity premium is a payment for a specific privilege — owning something few others can — and that privilege is financed by accepting a harder, slower, less certain exit. Price that privilege honestly and small editions are perfectly rational purchases. Assume scarcity is a free upgrade and you will overpay for illiquidity you never counted.
The band-by-band takeaways
Pulling the analysis back to the four edition bands, here is how each behaves for a buyer, and what to demand before you pay.
Editions of 150 or fewer
Ten prints in our set. Median price $231, but the widest liquidity range — some, like THESE SUNSETS (LARGE FORMAT) at 43 sales, trade well; others sit in the low single digits, where AK-47 Lotus (2 sales) and the smaller Sedation In Bloom variants live. This band is where scarcity can pay, but it is also where sample sizes get dangerous. Buy here for the object and the exclusivity, insist on strong provenance because you will not have a deep comp set to lean on when you sell, and accept in advance that your exit may take time.
Editions of 151–350
The sweet spot on price. Ten prints, the highest band median at $288, and healthy average liquidity of 53 sales. This is where marquee images at mid-size editions — Peace Goddess, MAKE ART NOT WAR (176 sales), Celebration Day (89 sales), Lenin Record — combine enough relative scarcity to hold price with enough volume to trade. If you want the scarcity premium without stepping into the liquidity desert, this band is the most defensible place to buy. It is also where the vintage prints cluster, which is likely why its median leads the table.
Editions of 351–550
The engine room. Twenty-one prints — nearly half the set — a median of $200, and by far the deepest liquidity at 77 sales per print and 1,611 total tracked sales. This is Fairey's workhorse band: the 400s, 450s, and 500s he releases most often. Prices are moderate, but price discovery is excellent and exits are easy. It also contains the two extremes — RBG's $460 and the sub-$180 letterpress and stencil runs — so the band average hides real spread. This is the band where an "edition of 450" tells you almost nothing about price by itself; you must look at subject, format, and demand. Buy here for liquidity and for the confidence that comes from a thick comp trail.
Editions of 551 and above
Seven prints, the lowest band median at $190, and solid liquidity at 59 sales per print. Large editions of Kurt Cobain (167 sales), Universal Dignity (149), and Snoop D-O Double G (112) trade constantly and cheaply. These are the accessible, liquid entry points to the artist — rarely where appreciation lives, but reliably where you can get in and out at a known price. For a first Fairey, or for a buyer who prizes an easy exit over a headline price, this band is underrated. Our companion piece on entry-level Fairey prints under $300 lives largely in this territory.
What the missing data won't let us claim
Honesty about the boundaries of this dataset is part of the analysis. Four limits matter.
First, this is 48 prints, not the full catalog. Our structural data counts 1,004 catalogued Fairey works and 595 with pricing, but only 48 in the clean comp set carry a verified edition size and release year together. The 595 per-file median prices are deliberately excluded here — they are too noisy to bucket by edition, and folding them in would launder guesswork into a table. So every band figure describes the 48-print comp set, a curated slice of the market's most-traded pieces, not a census. It is the best evidence we have; it is not the whole market, and a differently drawn sample could shift a band median by tens of dollars.
Second, some prints carry thin samples. Lenin Record (16 sales), Moon Over Biloxi (15), Basquiat Canvas (19), and several small-run canvases sit on comp counts where a single unusual result moves the average. We lean on medians over averages for exactly these prints, and even the medians deserve a mental error bar. Where a print appears with a low sale count, treat its price as indicative, not settled — the whole point of the liquidity discussion is that these numbers are softer than they look.
Third, edition size is silent on subject and format, and those often matter more — as the canvas and vintage cross-cuts showed directly. Our numbers capture the outcome but not always the cause, so read the edition band as one input into value, never the whole valuation. The comp valuation framework later in this series builds the full checklist that puts edition size in its proper, subordinate place.
Fourth, dated price-history points are individual transactions, not the median, and they scatter. A single $1,368 Peace Goddess result or a $125 Lenin Record result is a data point, not a market level. We cite them to show range and direction, never to set a price. When a headline number and a median disagree, trust the median.
How to use this when you buy
Turn the analysis into a decision rule. When a seller leans on scarcity to justify a price, run three checks against the comp set before you accept it.
- Is the edition genuinely small, or just "limited"? In Fairey's world, 300 is mid-size and 500 is large. An "edition of 450" is not scarce — it sits in the most liquid, most moderately priced band in the entire set. Do not pay a scarcity premium for a workhorse edition. Reserve that premium for the sub-150 band, and even there, verify the liquidity before you commit.
- Does the liquidity support the price? A high asking price on a print with a dozen comps is a price waiting to be tested. A high asking price on a print with 100-plus comps — RBG-style — is a price the market has already ratified hundreds of times. Deep liquidity de-risks a rich price; thin liquidity magnifies its risk. Ask how many comparable sales exist before you accept any number.
- Is scarcity or demand doing the work? If the print is expensive because few exist, you are buying illiquidity and must budget for a slow exit. If it is expensive because everyone wants it — even at a large edition — you are buying a liquid asset with a strong subject. The second is almost always the better financial position, and vintage plus a marquee image is the profile that has re-rated most reliably in our data.
The disciplined takeaway from 48 prints: scarcity is a lever, not a guarantee. It reliably raises unit price by roughly a third and just as reliably lowers liquidity by nearly half, and it ranks behind both demand and vintage as a driver of the number that actually clears. In a Fairey market with no external authenticator, liquidity is the quality most worth protecting, because turnover is the market's substitute for a certificate. The highest-conviction buys are the ones that command a strong price on deep volume — and those, more often than not, are driven by demand and age rather than by a small number stamped in the corner. Browse the full range on our Shepard Fairey collection, cross-reference any piece against the Fairey index, and for the scarcity-premium tier specifically, read the companion analyses on Fairey's highest-value prints and on what signed & numbered provenance adds to value.
This is analysis, not financial advice
Everything above is a data-driven reading of a 48-print comp sample drawn from public secondary-market sources (WorthPoint, eBay, LiveAuctioneers, Heritage), using a $90 signed-and-numbered floor and a rolling 2016–2026 comp window with trimmed-mean averaging. Art is an illiquid, subjective asset; past sale prices do not predict future results, and thin-sample prints in particular carry real estimation error. Nothing here is a recommendation to buy or sell any specific print. Do your own diligence, verify signatures, numbering, provenance, and condition in person, and treat these figures as a starting point for judgment, not a substitute for it.
