Introduction: Academic history has profoundly changed over the last five decades. The traditional subfields of 19th-century university degree courses—diplomatic, military, and political history—as well as the fields that rose to prominence in the 20th century, notably economic history, have been to a significant extent displaced by the histories of race and gender. The philosophy of history has all but vanished. Student protesters in the 1960s may have worried that their academic careers were in jeopardy after they challenged the relatively conservative faculty and administrators of their day. But today progressive faculty are dominant and enforce their political views to the point that the term “conservative historian” is close to being an oxymoron in most departments of history. These two phenomena are related. Academia’s ideological move to the left has undoubtedly changed which questions academic historians deem worthy of study, and which methods they judge to be appropriate. That the historical profession has radically changed is often discussed and—in conservative media—lamented. Yet very few historians have systematically explored the history of History, outside a few studies based on surveys spanning a few departments, subdisciplines, or decades. Few historians would dispute that there have been significant changes in the field over the past half century. Yet we lack a comprehensive view of the ways in which our discipline has changed, and how that change unfolded.
This essay has two goals. First, it aims to make the case for computational history, a methodological synthesis that combines classic historical approaches with the latest artificial intelligence (AI) tools. Second, it seeks to offer the first comprehensive and empirical exploration of the changes that have taken place in American university history departments from the 1870s, when the German university system rose to dominance in the United States, to today—just under 150 years.
Abstract: Fears of a second Gilded Age—of the excessive wealth and power of America’s biggest corporations—have put antitrust back on the Presidential agenda. On July 9, 2021, President Joe Biden released his “Executive Order on Promoting Competition in the American Economy,” which argued that “in the early 1900s, Teddy Roosevelt’s Administration broke up the trusts controlling the economy—Standard Oil, J.P. Morgan’s railroads, and others—giving the little guy a fighting chance.” The historical evidence shows otherwise. Federal intervention in leadingedge industries has followed media-driven shifts in public opinion, and the politics of antitrust as a “popular movement” produce mixed results for market competition—in part because activist leaders more often than not “could not tell the difference between federal regulation of business and federal regulation for business.” In theory, antitrust seeks to elevate market competition as the means to govern the economy. But in practice, antitrust action is historically accompanied by other types of regulation that have overwhelmingly benefitted incumbent businesses. Thus, in the context of government regulation, public outrage, and regulatory capture, antitrust will most likely reduce rather than increase competition, efficiency, and innovation in the American economy.
Although pandemics are not rare in history, societies rarely remember history’s lessons for containing disease. Using network science and the history of the last two great bubonic plague pandemics in 1340s and 1890s, including a detailed comparison of the fate of Renaissance-era Italian city-states, this paper traces the evolution of various disease-fighting policies and evaluates their effectiveness. Policies are most effective when they disrupt disease transmission networks by erecting borders between communities—maritime quarantines, sanitary cordons, and blockades. In contrast, policies that target a pathogen’s geography have sizeable economic costs and often produce social unrest—public hygiene and sanitation, housing changes, lockdowns, and policing. The paper concludes that the epidemiological models used to guide COVID-19 policy have inappropriately treated populations as homogeneous. Instead, whether Yersinia Pestis or COVID-19, policymakers should focus on understanding disease transmission networks—and how to disrupt them.
Abstract: Liquid staking bifurcates Ethereum’s validation into distinct economic and validation contributions. While reducing entry barriers for token staking, this approach consolidates validation within a limited number of specialized node providers. This paper outlines Heroglyphs, a protocol whose use creates incentives for “Complete Validators,” which are entities that are both economic and validation contributors, by refining Graffiti, a small piece of arbitrary data that validators can include in the blocks they propose. Heroglyphs consists of an Encoder for densely embedding information in Graffiti and a Translator for transforming this information into a variety of onchain operations, including but not limited to the creation, emission, transfer, and transformation of various tokens. Heroglyphs token mining ensures that all Complete Validators receive rewards for their participation, regardless of whether they are selected as block proposers and irrespective of their stake size, providing smaller validators with a stronger economic foundation. While we believe Heroglyphs will have applications far beyond the mining of tokens, for now Heroglyphs token mining returns fair token distribution mechanisms, pioneered by Bitcoin’s Proof-of-Work, to an Ethereum that has transitioned to Proof-Of-Stake.
Abstract: Buttonwood Poolside is a simple constant-product automated market maker (AMM) protocol optimized for use with value-accruing or rebasing tokens—VAR tokens for short. The most important VAR tokens to date include liquid staking derivative tokens (LSDs), real-world assets (RWA), bonds, and synthetic commodities like Ampleforth (AMPL). Liquidity providers (LPs) of all VAR tokens, whether rebasing or non-rebasing, incur avoidable divergence losses. For example, for native network assets and their corresponding LSDs, LPs will lose the yield from their LSDs. As of writing, LSD protocols have provided between $50M to $150M in yearly incentives to LPs in order to maintain liquidity for their tokens, which is crucial for broader integrations in decentralized finance (DeFi). Poolside mitigates these losses through the use of “reservoirs.” These are pockets of inactive liquidity that modulate changes in the value of VAR tokens. Additionally, as reservoirs grow in size, LPs can opt to deposit more of the matching, scarcer token so that both tokens can flow back into the active liquidity pool. To prevent reservoir manipulation attempts, the reservoir-matching function has three guardrails: a flow limit, a volatility circuit-breaker, and an approximated timeweighted-average-price (TWAP) requirement.
Abstract: Margin leverage is best suited for sophisticated traders. It is an instrument with asymmetric downside exposure which requires great risk management in volatile environments. All Decentralized Finance (DeFi) leverage today is margin leverage. To provide long-term permissionless credit, DeFi needs debt with zero margin calls and zero liquidations. In this paper we propose a design for one such debt instrument, which we call Buttonwood Zero, and which is a router contract built around the Buttonwood Tranche core contracts. We begin by offering a framework for understanding the effects of tokenizing debtor obligations and creditor claims. This allows us to classify various types of debt as zero-token, single-token, or double-token instruments. These insights frame our design for Zero, and how it can help anchor a web3 bond market.
Abstract: ButtonWood is a family of simple DeFi smart contracts whose primitive functions can be combined into more complex financial instruments. This short paper lays out one idea for how to combine two of these protocols into a “stablecoin” which we call “buttonStable.” The design borrows from financial history, and issues “derivative money notes” against a pool of “safe asset” debts, which are themselves the product of a volatility-tranching protocol called buttonTranche. The advantages of our design is that both the yield and leverage assets produced by buttonTranche do not need vaults or liquidation markets, so they can be freely traded. This makes it easy to price and manage risk. The process for minting a derivative money note is likewise simple, so buttonStable is both modular and robust.
Note: Tranche was previously known as “Button Alchemy”, a name abandoned after the Alchemy platform launch. This paper is revised to reflect the new naming terminology.
Abstract: Risk cannot be created or destroyed, it can only be transmuted, precipitated, and stratified. ButtonTranche (bTranche) is a protocol that, in recognition of this law, seeks to stratify risk in DeFi. The reactant is any rebasing crypto asset, and the products are “safer assets” and a “leverage asset.” Holding both these products recreates the risk profile of the original rebasing collateral. The protocol creates rolling zero-coupon bond tranches. Each “bond” (trancheBond) is initialized with a collateral and tranche ratio, and each TrancheBond contract mints and burns its own tranche tokens. Other users can then join the bond, minting more tranche tokens by adding collateral. Tranche tokens have varying volatility exposure, which is a product of their ordinal claims on the value of the locked collateral. Tranches A through Y at most trade at their par value, and are shielded from volatlity. The high-risk asset is the Z-tranche. The bTranche contracts are robust if the collateral is a rebasing asset. AMPL is a natural candidate, and so is a preferred choice. Other possible candidates can be created by using a continuously rebasing wrapper called buttonTokens, which is being built as part of the ButtonWood protocol family. TrancheBonds can be created with three main types of redemption mechanism: with a maturity date, without maturity but with constant redemption ratio, without maturity and without a redemption ratio. For reasons discussed below, our implementation requires redemption ratios, but can produce perpetual as well as maturing tranches.